Safety and Trespass Prevention - UIC - International … · ... Safety and Trespass Prevention PROCEEDINGS 11/208 Session 1 - Risk Management Chairman: Luc Bourdon, Transport Canada,

www.levelcrossing2008.com

Organised by:

In cooperation with:

1100tthh WWoorrlldd LLeevveell CCrroossssiinngg SSyymmppoossiiuumm SSaaffeettyy aanndd TTrreessppaassss PPrreevveennttiioonn

PPRROOCCEEEEDDIINNGGSS

2244tthh

–– 2266tthh

JJuunnee,, 22000088 UUIICC HHeeaaddqquuaarrtteerrss,, PPaarriiss

www.levelcrossing2008.com

10th World Level Crossing Symposium10th World Level Crossing SymposiumSafety and Trespass PreventionSafety and Trespass Prevention


PROCEEDINGS 3/208

Contents

AAcckknnoowwlleeddggeemmeennttss ................................................................................................................................................................................ 66

SSyymmppoossiiuumm SScchheedduullee .......................................................................................................................................................................... 77

VVeennuuee .............................................................................................................................................................................................................................. 77

SSppoonnssoorrss .................................................................................................................................................................................................................. 88

EExxhhiibbii ttoorrss ................................................................................................................................................................................................................ 88

PPRROOCCEEEEDDIINNGGSS .......................................................................................................................................................................................... 1100

SSeessssiioonn 11 -- RRiisskk MMaannaaggeemmeenntt ............................................................................................................................................ 1111

Impact of Traffic Flow on Level Crossing Risk .................................................................................................11 Level Crossing Safety Performance Monitoring by Web based Knowledge Management System ................19 Development, Implementation and Use of the All Level Crossing Risk Model ................................................33


Recent experience in Victoria, Australia ..........................................................................................................40 Statistical analysis of risk .................................................................................................................................47 French Safety Improvement policy of Level Crossings ....................................................................................53


Level Crossing Accidents in Indian Railways – Risk Reduction Measures......................................................54 Safety inspections at Finnish level crossings...................................................................................................63


Proposition of a New Level Crossing Protection System.................................................................................64 A Risk Assessment framework for Road-Rail Level Crossings: Application to Moroccan Level Crossings ....65 Corporate responsibility for pedestrian risks at level crossings .......................................................................74

SSeessssiioonn 55 -- SSaaffeettyy aanndd HHuummaann IIssssuueess .................................................................................................................. 7755

JR East’s Efforts in Level Crossing Safety.......................................................................................................75 Why do people trespass? Finnish Experiences ...............................................................................................81 Current Status of Level Crossing Accidents and Solutions for Enhancing LC Safety in Japan.......................88

SSeessssiioonn 66 -- SSaaffeettyy aanndd HHuummaann IIssssuueess .................................................................................................................. 9988

Operation Lifesaver - Reducing Level Crossing Collisions and Trespassing Incidents through Education ....98 Rail Safety Education Strategies in the United States ...................................................................................105 A new educational approach for preventing Level Crossing Accidents .........................................................111 Innovative and Cooperative Leadership to Improve Safety at Australian Level Crossings ...........................118

SSeessssiioonn 77 -- EEnnggiinneeeerr iinngg aanndd OOppeerraatt iioonn.......................................................................................................... 112266

Vision Zero – Guiding safer level crossings ...................................................................................................126



PROCEEDINGS 4/208

A Technology Comparison of Two In-Vehicle Warning Methods at Level Crossings with Human Factor Implications.....................................................................................................................................................137 Risk assessment – is it a best possible method for LC categorisation? ........................................................144


The impact of changing legal powers for public road level crossing regulation in Great Britain, and use to influence better partnership working. .............................................................................................................151 Possible Measures for Safety Increase of ŽSR Level Crossings...................................................................160 Bentleigh, Victoria – Australia - a Railway Pedestrian Crossing Case Study ................................................167


Customised techniques and operational rules to improve level crossings by means of imaging methods ..176 Site Research, Simulation and Evaluation of Novice Engineering Safety Devices at Israeli Level Crossings........................................................................................................................................................................184 The railway is not an island: Building partnerships with the wider community ..............................................185

SSeessssiioonn 1100 -- FFuuttuurree VViissiioonn .................................................................................................................................................. 119944

Australian Road/Rail Strategy – development of the joint national interface strategy in Australia ................194 Applying a partnership approach to level crossing risk: a strategic opportunity ............................................195 Moving forward together.................................................................................................................................202

TTeecchhnniiccaall VViissii tt .......................................................................................................................................................................................... 220033

Highway Code ................................................................................................................................................203 Presentation of the Breuillet level crossing number 30..................................................................................205 Presentation of the planned removal operation .............................................................................................206



PROCEEDINGS 5/208

Organiser

International Union of Railways

In Cooperation with

Réseau Ferré de France

Level Crossing Organisation Team Mr. Simon Fletcher Ms. Isabelle De Keyzer Mrs. Isabelle Fonverne Mr. Jérôme Prieur

Level Crossing Review Team Mr. Simon Fletcher, UIC Mr. Philippe Feltz, RFF (Réseau Ferré de France), France Mr. Yves Mortureux, SNCF (Société Nationale des Chemins de Fer Français), France Mrs. Kirsi Pajunen, Finnish Rail Safety Authority, Finland Mr. Lazlo Tordai, UIC Mr. Michael Robson, Secretary General EIM (European Rail Infrastructure Managers), Mr. Wallace Weatherill, South Eastern Ltd, United Kingdom Mr. Michael Woods, RSSB (Rail Safety and Standards Board), United Kingdom Mr. Aidan Nelson, Community Safety Partnerships, United Kingdom Mr. Witold Olpinski, CNTK (Railway Scientific and Technical Centre), Poland Mr. Alan Davies, RSSB (Rail Safety and Standards Board), United Kingdom Mr. Peter Gerhardt, DB (Deutsche Bahn), Germany Mr. Jerôme Prieur, UIC



PROCEEDINGS 6/208

Acknowledgements The 10th World Level Crossing Symposium was only the second time that this event has taken place in Europe and we were delighted to have had the opportunity to organise the event. The road/rail interface is a very complex subject, clearly demonstrated by the enthusiasm with which debates and round tables were conducted in response to the extremely high quality of papers that were presented. 166 delegates from 37 different countries attended and there were over 30 speakers and a number of invited guests from organisations such as the European Commission, the European Railway Agency and other European institutions. With the event being in Europe much of the focus was on the road/rail interface there and it was with considerable interest that the delegates were able to learn about the experiences of programmes as far away as USA, Canada and Australia where educational and behavioural studies form a great part of the work in this area. Much discussion is going on in Europe about next steps and the desire for the European Level Crossing Research Forum (which was formed out of the 8th Level Crossing Symposium in Sheffield in 2004) to take a lead role in developing the proposed Road/Rail Interface Strategy at European level. This would see the engagement on a regular basis with the other sectors playing a key role in the road/rail interface – law enforcement, government agencies etc., as well as other sectors that can help to build a greater awareness around the management of the risks at this interface. We would like to express a number of thanks for those who helped the organising team to make this event such a success – speakers, delegates, guests, the team at the UIC… as well as the exhibitors who so enthusiastically demonstrated their products and services and to Thales who helped to sponsor the event. These proceedings are a collection of all the papers that were presented to the Symposium and have been put together jointly by Jérôme Prieur, and Isabelle De Keyzer at UIC who were also responsible for the majority of the organisation of the event. The road/rail interface is one that continues to cause the rail sector a range of problems and which can only be overcome through sustained effort and a partnership approach with the other sectors. This is a role that we can all contribute to and I look forward especially to working with you all as the UIC supports the rail community in developing our alliances with those other sectors. I look forward to seeing you in Tokyo in 2010. Kind regards,

Simon Fletcher on behalf of the Organising Team



PROCEEDINGS 7/208

Symposium Schedule

Venue The Level Crossing 2008 was held in Paris at the UIC headquarters near the Eiffel tower. UNION INTERNATIONALE DES CHEMINS DE FER 16 rue Jean Rey 75015 Paris

Tel: +33 (0) 1 44 49 20 20 Fax: +33 (0) 1 44 49 20 29

Access � Métro : Line 6, Bir Hakeim � RER : Line C, Champ de Mars � Bus : lines 42, 69, 82, 87 � Taxi : Easily available

Monday 23 June

Tuesday 24 June

Wednesday 25 June

Thursday 26 June

Friday 27 June

09:00 – 15:00 Technical Visits

09:00 – 17:30 Symposium

09:00 – 17:30 Selcat Meeting

09:00 – 17:30 Symposium

16:45 - Poster Session

09:00 – 17:30 Symposium

18:00 - Welcome Cocktail Free Evening 19:15 - Official dinner Free Evening

End of Symposium



PROCEEDINGS 8/208

Sponsors

THALES supported the Symposium as Silver sponsor. With its global network of 22,000 high-level researchers and operations in 50 countries and 68,000 employees, Thales is a world leader in Mission-critical information systems for the Aerospace, Defence, Security and Transport markets. THALES RAIL SIGNALLING SOLUTION, S.L.U c/. Serrano Galvache, 56 Edificio Alamo – 4 Planta Norte Madrid 28033 Spain

T: (3491) 273 7534 E: [email protected] www.thalesgroup.com

Exhibitors A small exhibition was set up with stands from the following companies:

Pintsch BAMAG Product and services description: PINTSH BAMAG stands for highest degree of reliability and competence as the supplier of complete level crossing protection systems for mainlines as well as for private, industrial, and port railways. The failsafe systems are characterised by their simple operability and high availability, at low investment and lifetime costs. PINTSCH BAMAG Hünxer Straße 149 Dinslaken 46537 Deutchland

T: (0049) 2064 602 370 E: [email protected] www.pintschbamag.de



PROCEEDINGS 9/208

Eldor Communication Technologies Ltd.

Product and services description: Eldor Communications Technologies Ltd. offers reputable system integrator with expertise in video analytics based systems. Eldor has developed an advanced detection and warning system of potential collision hazard in the Level Crossing area called Collision Avoidance & Safety Alert (C.A.S.A. ). This system includes all advanced (video, thermal, inductive) detection elements and processing (video analytics) required to supply alarm warning when a vehicle is obstructing the Level Crossing area and constitutes a potential dangerous hazard. The warning is provided to the RR signalization system, to the RR Control Station and to the train driver (optional) via WAN and wireless communication infrastructure. Eldor Communication Technologies Ltd. 22 Hamelacha St. PARK CIBLE P.O.B 11401 Rosh Haayn 48091 Israel

T: (972) 390 247 50 E: [email protected] www.eldor-il.com

Strategic Thought Limited

Product and services description: Strategic Thought specialises in Risk computer software and services. In conjunction with the Network Rail, the UK’s railway network operator, Strategic Thought developed the All Level Crossing Risk Model (ALCRM). ALCRM is a software solution to predict Risk at rail crossings. ALCRM can help rail network and operating companies to:

- Assess the safety of all level and rail crossings, from automatic crossings, manually controlled gate crossings to footpaths

- Allocate resources effectively to prevent accidents based on the assessed level of risk Visit the Strategic Thought stand to take a closer look at ALCRM or to discuss your risk management requirements in general. Strategic Thought Group The Old Town Hall 4 Queens Road London, UK SW19 8YA

T: +44 (0) 20 8410 4000 E: [email protected] www.strategicthought.com



PROCEEDINGS 10/208

PPRROOCCEEEEDDIINNGGSS



PROCEEDINGS 11/208

Session 1 - Risk Management Chairman: Luc Bourdon, Transport Canada, Canada

Impact of Traffic Flow on Level Crossing Risk

Author(s): Jay Heavisides, Dr John Barker

Job Title: Safety & Risk Consultant

Company: Arthur D. Little

Country: United Kingdom

Resume of Speaker

Jay Heavisides is a Manager at Arthur D. Little, where she has worked on a variety of assignments in both the transportation and process industries over the last seven years. Specializing in safety and risk management, her experience covers a range of areas including quantitative risk analysis, consequence and safety/cost benefit model development, technical evaluations, audit, and safety case development. Most recently she has been involved in conducting reviews of level crossings risk factors in three different European countries and research into the benefits of “another train coming” warnings. Ms. Heavisides has a Masters degree in Chemical Engineering from Imperial College, London, is a Chartered Engineer and is a Member of the Institute of Chemical Engineers.

Abstract

This paper considers the assessment of traffic flow as a factor determining risk at level crossings. It builds on theories first introduced by Professor P. F. Stott on the safety of automatic open crossings. Stott proposes that driver behaviour is conditioned less by the level crossing protection, and more by the vehicle ahead. This means that at busy crossings relatively few motorists actually have the opportunity to respond to the crossing equipment, instead reacting to a stopped vehicle ahead that has arrived first at the crossing. The opportunity for a collision with a train may therefore actually decrease at higher traffic flows. These theories are supported by anecdotal evidence of accident patterns at Automatic crossings in Great Britain; in recent years some of the more serious train-vehicle collisions have been at Automatic crossings with only moderate traffic flows.

Introduction

Evaluating risk at level crossings to understand where to focus investment in risk mitigation is now common practice in many countries [1]. Utilisation is considered to be a common driver of collision opportunity. One way to measure utilisation and its impact on risk is to consider Traffic Moment, as given in Equation (1).



PROCEEDINGS 12/208

Traffic Moment = Number of trains x Number of users (1)

The equation suggests that, as the number of trains or the number of users increases at a crossing, so does the opportunity for collision (see Figure 1); i.e. it is a linear model. Use of Traffic Moment to model exposure to collision risk makes an important assumption: that trains and users arrive at the crossing independently and that each train and user has an equal opportunity for collision.

0

1

2

3

4

5

6

7

8

9

0 1 2 3 4 5 6 7 8 9

Traffic Moment (per day)

Opp

ortu

nity

for

colli

sion

Figure 1: Linear model for opportunity for collision

However, work by Professor P. F. Stott [3] at Automatic Open Crossings (AOCs) suggests that not all vehicle users have an equal opportunity for collision. Stott proposes that driver behaviour is conditioned less by the level crossing protection, and more by the vehicle ahead. This means that at busy crossings relatively few motorists actually have the opportunity to respond to the crossing equipment, instead reacting to a stopped vehicle ahead that has arrived first at the crossing. Stott argues that, for a road vehicle-train collision to occur at an AOC a vehicle must arrive at the crossing just before the train and fail to stop1. The conditions leading to a collision are not always met, for example:

• If the vehicle arrives immediately before the train and stops

• If the vehicle arrives significantly before the train, fails to obey the warnings and does not stop

• If a vehicle arrives significantly before the train and stops, the next vehicle then arrives just before the train but on seeing the stationary car at the crossing also decides to stop

In the last example the first vehicle to arrive acts as a barrier to the second vehicle, thus reducing the opportunity for collision. At crossings with low traffic flows a car arriving just before the train is likely to be the first – giving a greater potential for collision per vehicle. The driver here must decide whether or not it is safe to proceed over the level crossing. At crossings with high traffic flows, a car arriving just before the train is not likely to be the first; emphasis therefore shifts from reacting to the crossing equipment to reacting to the stationary car, reducing potential for collision. This means that, at AOCs, the opportunity for collision from red light running is not proportional to the number of road vehicles but is instead non-linear (as illustrated in Figure 2).

1 Stott considers ‘failing to stop’ as continuing over the crossing past the protecting red lights deliberately or by

accident



PROCEEDINGS 13/208

Mean flow (vehicles/second)

Opp

ortu

nitie

s fo

r th

e 1s

t veh

icle

arr

ivin

g ju

st b

efor

e th

e tr

ain

Lower traffic levels: Opportunities for collision are approximately linear to the increase in traffic flow

Moderate traffic levels:Queuing starts to occur at the crossing, providing a protective barrier - reducing the proportion of vehicles that can arrive directly at the crossing just before the train

Higher traffic levels:Queuing frequently occurs at the crossing, providing further layers of protection against vehicles arriving directly at the crossing just before the train and reducing the overall opportunity for collision

Figure 2: Stott’s non-linear collision model at AOCs

Assuming that road vehicles arrive at random, Stott [3] went on to model the non-linear opportunity for collision using a Poisson distribution (as shown in Figure 2). This is expressed in the following equation:

PC = exp(-M(T-t))-exp(-MT) (2)

Where:

PC is the probability of that the first vehicle arrives at the crossing just before the train

M is the mean traffic flow over the crossing (vehicles/second)

T is the strike-in time (the warning provided by the crossing that the train is approaching, in seconds)

t is the time interval the vehicle must arrive in before the train to result in a collision

Stott used the Poisson distribution to argue that a crossing with a moderate traffic flow presents the greatest opportunity for collision – the point at which traffic flow is providing greatest opportunity for vehicles to arrive just before the train at the level crossing. This is contrary to the linear model, which suggests that the busiest crossing has the most opportunities for collision.

Application of Stott’s theory at other crossing typ es

Stott [3] developed the non-linear model with AOCs in mind. However, the non-linear model is also relevant to other crossing types where traffic normally flows over the crossing without stopping.

Automatic half-barrier (AHB) crossings have additional visible protection through the provision of the barrier that closes the entrance to the crossing (the barrier itself providing limited physical protection of the crossing from any approaching vehicle). At AHBs a driver may deliberately choose to disobey the warnings and zigzag around the protecting barrier. If the zigzag occurs a few seconds before arrival of the train this could lead to a



PROCEEDINGS 14/208

collision. This behaviour is similar to the red light running at AOCs considered by Stott, and can therefore also be modelled using Equation (2). Both AOC and AHB crossings are initiated automatically (as suggested by their name) as the train passes over a designated point on the approaching track. In Great Britain AOC and AHB crossings are designed to provide a user with a minimum warning time of 27 seconds for any approaching train; 50% of trains should arrive within 50 seconds and 95% of trains within 75 seconds [2]. Across the British network the average warning time at AOC and AHB crossings is around 34 seconds [4]. Short crossing closure times are justified as the exit from these types of crossing is always open, allowing any user entering the crossing as it is closing an opportunity to escape. Traffic flow at both AOC and AHB crossings are thus affected by similar crossing closure times.

At full barrier crossings (in Great Britain these are manually operated) a driver is unlikely to deliberately drive through the barriers (probably in fear of damaging their vehicle) – this is supported by accident history: no road vehicle-train collisions have been recorded in the last 15 years at full barrier crossings in Great Britain.

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

0 0.05 0.1 0.15 0.2 0.25 0.3

Mean traffic flow (vehicles/sec)

Opp

ortu

nity

for

colli

sion

27 sec

34 sec

50 sec

75 secAs warning time increases the probability of the first car arriving just before the train reduces

However, longer warning times are more likely to result in users deliberately running or zigzagging the crossing

Warning time

Figure 3: Effect of warning times on opportunity for collision

As the warning time increases the overall probability of the first car arriving just before the train decreases – the first car is more likely to arrive outside this critical window. However, it should be noted that this model does not account for driver impatience; longer warning times at AOC or AHB crossings are more likely to result in users deliberately running the red lights or zigzagging the barriers. Therefore in practice there is a trade-off in the overall probability of collision between warning times and user impatience.

Application of Stott’s theory for other causes of r oad vehicle-train collisions

Not everyone involved in a collision at AOC or AHB crossings runs the red lights or zigzags. For example a vehicle driver could:

• Not be aware of the crossing because of fog, a bend in the road obscuring the approach, low angle of the sun etc.

• Misjudge the braking distance because of poor road surface, the approach is at a downward gradient, the presence of mud or ice etc.



PROCEEDINGS 15/208

• Stop safely in front of the crossing but then struck from the rear by another road vehicle shunted into the path of the train

Traffic flow also influences the number of opportunities for road vehicle-train collisions when driver error occurs.

In the event of the driver not being aware of the crossing, the vehicle would have to arrive at the crossing just before the train for there to be an opportunity for a collision. Subsequent vehicles are unlikely to be involved in a collision through driver error because the first vehicle both acts as a physical barrier to the crossing and a visible warning that there is a hazard ahead. The influence of driver error on opportunities for collision is therefore similar to red light running and therefore can be modelled using Equation (2).

In the event of the driver misjudging the required safe braking distance and stopping on the crossing, there is a greater opportunity to be involved in a collision with the train as the vehicle remains placed in front of the approaching train until the driver takes further action. To avert the collision the driver must therefore have sufficient time between stopping and the train arriving to be able to recover. Subsequent vehicles are again unlikely to be involved in a collision through driver error because the first vehicle both acts as a physical barrier to the crossing and a visible warning that there is a hazard ahead.

Accidents of this type could also occur at full barrier crossings, although it should be noted that none have been reported to have resulted in collision with a train.

When considering the opportunity for collision, the key difference between a driver misjudging the braking distance and a driver being unaware of the crossing is therefore the time window for the causal event to occur. This window is much greater for a collision caused by misjudging the braking distance than one caused by being unaware of the crossing. For there to be an opportunity for collision the first car that arrives at the crossing must misjudge the braking distance within this time window (taken to be the warning time) and be unable to recover prior to arrival of the train.

As the warning time increases so does the opportunity for a first vehicle to arrive at the crossing a head of the train. All road vehicles have potential to misjudged braking distance and stop on the track. This means that the associated opportunities for collision tend to reach the ‘peak’ towards lower traffic densities (see Figure 4).

0%

20%

40%

60%

80%

100%

0 0.05 0.1 0.15 0.2

Mean traffic flow (vehicles/second)

Opp

ortu

nitie

s fo

r co

llisi

on w

ith tr

ain

from

m

isju

dgin

g br

akin

g di

stan

ce

27 sec

34 sec

50 sec

75 sec

180 sec

However, longer warning times increase the opportunity for the driver to recover

Longer warning times give greater opportunity for a vehicle to brake late and come to rest on the crossing in the path of an approaching train

Warning time

Figure 4: Misjudging braking distance collision model



PROCEEDINGS 16/208

Using the Poisson distribution the opportunity for collision from misjudging braking distance can be modelled using Equation (3).

PC = 1 – exp(-MT) (3)

By increasing the warning time or crossing closure time (denoted in (3) by T) it is possible to see that the opportunity for collision also increases and for long warning times the opportunity for collision can reach its peak at low or moderate traffic flows.

For low traffic flows with long crossing closure times (e.g. a full-barrier crossing being closed for around 180 seconds prior to the arrival of the train) would present the greatest opportunity for a collision from a vehicle braking late. However, given that the warning of the approaching train is longer, it also provides the greatest opportunity for driver to recover (not included in (3)) and if the crossing is locally manned, the crossing keeper may be able to assist in moving the obstructing vehicle or slowing/stopping the train (the same cannot be said for full barrier crossings operated remotely by CCTV). Crossings with short warning times (such as AOC and AHB crossings) have reduced opportunity for a collision from misjudging the braking distances but provide a driver with the least opportunity to recover. The effect of a driver being able to recover is estimated in Figure 5.

0%

20%

40%

60%

80%

100%

0 0.05 0.1 0.15 0.2


Opp

ortu

nitie

s fo

r co

llisi

on w

ith tr

ain

from

m

isju

dgin

g br

akin

g di

stan

ce (

incl

udin

g dr

iver

re

cove

ry)

27 sec

34 sec

50 sec

75 sec

180 sec

Warning time

Figure 5: Misjudging braking distance and then recovering

Figure 5 shows that with driver recovery it is still possible for the opportunity for collision to reach its peak at moderately busy crossings.

Collisions caused by rear-end shunting require 2 cars to arrive at the crossing before the train: the first car stops and the second car hits the first into the crossing, which is then pushed onto the tracks. This is similar to misjudging braking distances as the window of opportunity is significantly greater than for red light running or being unaware of the crossing. As the traffic flow increases so does the potential for at least 2 cars to arrive at the crossing during activation, until there is always potential for 2 cars to arrive during a single activation (see Figure 6).



PROCEEDINGS 17/208

0%

20%

40%

60%

80%

100%

0 0.05 0.1 0.15 0.2 0.25


Opp

ortu

nitie

s fo

r co

llisi

on w

ith tr

ain

from

rea

r-en

d sh

untin

g

27 sec

34 sec

50 sec

75 sec

180 sec

Warning time

Figure 6: Rear-end shunting collision model

Assuming a Poisson distribution, the opportunities for rear-end shunting leading to train collisions can be modelled using Equation (4).

PC = 1 – exp(-MT) – MT exp(-MT) (4)

Again varying the warning/crossing closure time before the train arrives, it is possible to see that crossings with longer closure times have the greater opportunity for rear-end shunting accidents to occur at lower traffic flows: as the warning time increases so does the opportunity for 2 cars to arrive at the crossing. Similar to ‘misjudging the braking distance’: the longer the warning the greater the opportunity for the vehicle driver to recover.

Effects on traffic flow on risk modelling

The ‘non-linear’ models described above were adapted for implementation in Network Rail’s All Level Crossing Risk Model (ALCRM). The ALCRM is now being used in the risk assessment of all level crossings on Network Rail managed infrastructure. Previously in Great Britain a linear collision model [4] had been used to assess risks at automatic crossing. Comparison of results from both models shows some interesting differences. Of the top 5 highest-risk crossings modelled using the linear model; only 2 remain in the top 5 predicted to be highest-risk by the non-linear model. Of the 20 crossings predicted as highest-risk by the linear model, only 8 appear in the same set from non-linear risk model. The changes observed when using the non-linear model appear to agree with anecdotal evidence on the pattern of fatalities and serious injuries at these crossings.

Conclusions

Non-linear models of the opportunities for train-user vehicle collisions appear to correct much of the historic over-prediction of risk at GB’s busiest crossings. The models also show how opportunities for collision (and therefore total risk) may actually peak at crossings that are moderately busy. These theories are supported by anecdotal evidence of accident patterns at Automatic crossings in Great Britain; in recent years some of the more serious train-vehicle collisions have been at Automatic crossings with only moderate traffic flows.



PROCEEDINGS 18/208

The All Level Crossing Risk Model (ALCRM), developed by the Rail Safety & Standards Board (RSSB) and implemented by Network Rail [5], uses non-linear models similar to those described here in its assessment of level crossing risk. The output from the ALCRM will provide valuable information on where to best focus expenditure on risk mitigation.

Future research must concentrate on refining these models, particularly where human factors play a role in determining how crossing users respond to operation of the crossing and to other users.

Acknowledgements

The authors would like to acknowledge the Rail Safety & Standards Board (RSSB) for their support and funding of this and other related research, including:

• Documenting the All Level Crossings Risk Model (T737) • Research into the Safety Benefits Provided by Train Horns at Level Crossings (T668) • Examining the Benefits of ‘Another Train Coming’ at Automatic Level Crossings (T652) • Use by Other Railways of Risk Models and Risk Assessments for Level Crossings (T524) • Research into Obstacle Detection at Level Crossings (T522) • Developing Enhanced Consequence (Derailment) Algorithms for Level Crossing Risk Models (T521) • Cost of Level Crossings – an International Benchmarking Exercise (T364) • Understanding Risk at Station and Barrow Crossings (T332) • Development of Universal Level Crossing Risk Tool (T028) • User Worked and Footpath Level Crossing Research (T000)

Further details of this research can be obtained from www.rssb.co.uk

The opinions expressed in this paper are those of the authors.

References

[1] Beard, M. J., Chawla H., “Use of Risk Models and Risk Assessment for Level Crossings by Other Crossings”, Research report by Arthur D. Little on behalf of the Rail Safety & Standards Board, (2007).

[2] Health & Safety Executive “Railway Safety Principles and Guidance, part 2 section E Guidance on level crossings”, HSE Books

[3] Stott, P. F. “Automatic Open Level Crossings A Review of Safety”, Paper to the Department of Transport, (1987).

[4] “Automatic Level Crossing Risk Model v3.1”, Network Rail [5] “All Level Crossing Risk Model (ALCRM)”, Network Rail



PROCEEDINGS 19/208

Session 1 - Risk Management

Level Crossing Safety Performance Monitoring by Web based Knowledge Management System

Author(s): Dr Roman Slovak, A. G. Schielke, E. Schnieder

Job Title: Senior Research associate

Company: University of Braunschweig, Institute for Traffic Safety and

Automation Engineering

Country: Germany

Resume of Speaker

Dr.-Ing. Roman Slovák (born 1974) received the M.S. (in electrical engineering with specialization on railway safety systems) and the Ph.D. in mechanical engineering. Since 1999 he is working at the Institute of Traffic Safety and Automation Engineering at the Technical University of Braunschweig as a research assistant. His main research interests are in application of formal methods for methodical modelling and analysis of railway safety systems as well as in development of tools for support of railway operation control

Abstract

The paper presents the actual and planned functionality of the Level Crossing Knowledge Management System developed in the context of the European project SELCAT. The focus is given on the functions related to the monitoring of the level crossing safety performance. After a brief introduction in the objectives of SELCAT the main interfaces for data collection are outlined. Finally the paper presents the main features of the system allowing analysing the collected data on level crossing accidents.

Introduction

The Coordination Action "SELCAT" (Safer European Level Crossing Appraisal and Technology), is responding to the call of 6th Framework Programme of the European Commission in the area of "Sustainable Surface Transport Coordination Actions" towards the objective "Increasing road, rail and waterborne safety and avoiding traffic congestion". It aims actively to contribute to the reduction of level crossing accidents by the:

• collection, analysis and dissemination of existing research results and the stimulation of new knowledge exchange in the area of level crossing safety,

• creation of circumstances whereby European partners, in the rail and road sectors, can make a significant contribution to the reduction of accidents, injuries and fatalities at level crossings,

• understanding and codifying of existing and planned research,

• comparison and harmonisation of data sources,



PROCEEDINGS 20/208

• exploration of new technologies and harnessing appraisal techniques to optimise these.

The SELCAT consortium led by Institute for Traffic Safety and Automation Engineering of Technical University of Braunschweig, Germany, integrates 25 partners from 14 countries from Europe, Asia and Africa. Among them are universities, research institutes, road and railway organisations as well as railway infrastructure managers. The SELCAT consortium is closely collaborating with European Railway Agency (ERA).

The activities of SELCAT lead directly to the improvement and expansion of inter-modal collaboration between the road and rail sectors. The information collection, exchange and comparison is be provided by creation of a "Level Crossing Web Portal". It should result in an effective and internet accessible Level Crossing Knowledge Management System (KMS) allowing also broad dissemination of safety and level crossing related research activities investigated by the SELCAT Project.

The paper presents preliminary results of the Work package 1 dealing with the Level Crossing Appraisal. One of the main aims of this work package is to estimate the global risk of different types of level crossings. As input for this task the accident statistics of countries of involved partners were considered.

Motivation

Every year, more than 330 people are killed and several hundreds of injured in more than 1200 accidents at road-rail level crossings in the European Union. Together with tunnels and specific road black spots, level crossings have been identified as being a particular weak point in road infrastructure, seriously affecting road safety [1]. In the case of railway transport level crossings can represent as much as 50% of all fatalities occurring in railway operation. Up to now, the only effective avoiding appears to involve upgrading level crossing safety systems [2] even though in more then 90% of cases the primary accident cause is inadequate or improper human behaviour rather than any technical, rail-based issue.

The Safety Directive 2004/49 EC [4] as well as the CENELEC standards for railway application expects application of risk based approaches when eliminating the negative statistics of occurred accidents. Thus the evaluation of operational risks plays the crucial role in any kind of decision making regarding the change of operational rules, legislation or technical systems.

From these reasons in the context of the SELCAT project a structure for a safety performance database is to be designed and to be filled with data from available databases, especially from involved infrastructure managers. Attention should be paid to the integration of detailed information on level crossing accidents (collisions with cars, cyclists, pedestrians etc.), their causes (safety system, human factors, etc.) taking into account the safety equipment (none, lights, barrier, etc.) and operational condition of the rail (single/double line, main/local line, average traffic flow, etc) and road traffic (infrastructure conditions, traffic flow, sight conditions, pedestrians flow, etc).

The aim is to identify the level crossing types which cause the highest operational risk. This should be allowed to be carried out in general as well as taking into consideration the country specific operation condition including human behaviour, legislation, degree of technical development.

Approach

A detailed evaluation of the level crossing operational risk can be based only on the national statistics coming form different partners of the SELCAT project. The application of existing European statistics (e.g. EUROSTAT) has been shown as unusable as they do not provide the required level of information details. The same problem occurs when analysing the most official national statistics sources.

Any kind of statistics collection and comparison requires harmonisation of their sources. Therefore a challenging aim of SELCAT became to design an universal platform allowing integration of statistics from different partners related to different national specific level crossing types. The chosen approach starts with a



PROCEEDINGS 21/208

separation of abstract functions and concrete resources performing the functions. Hence the functionality as the basic structuring element was investigated principally.

Structuring of the level crossing functionality is based on a generic approach [5] [6]. A level crossing (LC) is generally seen as crossing of railway and road traffic flows (basic dynamic or functional LC aspects) which safety related physical interaction must be prohibited by the operation functions of the level crossing safety system (static or physical LC aspects) [7]. In order to describe the static LC aspects the generic approach identifies four basic operational functions (to detect, to inform, to warn and to protect) on both traffic sides (see figure 1). The basic functions are further refined e.g. the function “to warn” is further split according to its general way of realisation (audible, visual, physical).

Figure 1: Functional decomposition of a level crossing system

A step to the technological concretisation represents a further refinement of the basic structure of static operation aspects. The final refinement level is given by a particular technological solution given by the level crossing type and the way of operation (automatic, manual).

The dynamical aspects of the level crossing are covered by the definition of operation condition of the traffic flows. Here can be integrated also temporal aspects of the safety procedure of a level crossing.

Beside the static and dynamic aspects a major part of the domain knowledge concerns organisational and legal aspects which also include construction laws and standards. This includes also other lifecycle aspects.

Implementation of the Knowledge Management System

Level crossing types

The implemented interface of the SELCAT KMS allows to the project partners specifying all national level crossing types using predefined functional structure. On this way is guaranteed the common specification language, necessary for the future risk analysis.

Figure 2 shows an example of a defined and classified LC type. On this way up to now defined about 70 national specific level crossing types has been defined. Apart of identification of level crossing functions each type definition contains also data about the approximate acquisition and operation cost for the purposes of any future cost benefit analysis. Included are also relevant data of national legislative e.g. maximal and minimal allowed traffic flows, speeds on railway and road side. Finally the existing population of particular level crossing type is included allowing normalization of future analysis results. Beside the level crossing type population also some further general information for normalization of the particular country are collected (length of railway network, train kilometres, human population per square kilometre, total number of road or railway accidents, etc.)



PROCEEDINGS 22/208

The KMS provides the possibility to search interactively for level crossing types according to given functionality or technical implantation features. The different searching descriptors can be connected by AND or OR gate. Beside interactive search one is able to define general level crossing types (basic types) on a more abstract level of functionality definition. In the case of the abstract level crossing type the user indicates the relevance of functions for his analysis (existing, not existing and irrelevant functions). Such abstract (meta) level crossing type can be used for search of groups of related particular types from different countries. As an example of such abstract level crossing types the categories of European Railway Agency have been defined [8]. The figure 3 shows schematically this categorisation and the figure 4 an exemplar definition of the basic type A.1.3. The figure 5 shows results of the matching procedure where the collected national LC types has been investigated and matched with one of the defined the basic type (A.1.3).

Figure 2: Example of a LC type in the SELCAT KMS (source: Railway Safety & Standard Board (RSSB), UK)



PROCEEDINGS 23/208

Figure 3: Proposal of European Railway Agency for the level crossing classification

Figure 4: Example of a LC basic type definition in the SELCAT KMS (ERA Type A.1.3)



PROCEEDINGS 24/208

Figure 5: Example of searching for related level crossings to the base Type A 1.3

Accident statistics collection

The KMS allows to each defined Level crossing type to associate the related accident statistics. In particular it is possible to collected statistics about:

• the accident severity (fatalities, serious and light injuries), • the kind of the accident (car, bus, bicycle, pedestrian, etc) and • the accident causes (external, internal, technical, human).

The figure 6 shows an example of level crossing statistics referring to two different level crossing types for reference years 2001 - 2005.

Figure 6: Example of level crossing accident statistics entered in the SELCAT KMS: AHB – Automatic half barrier crossing, AOCL - Automatic open crossing locally monitored (source: RSSB, UK)



PROCEEDINGS 25/208

Accident statistics of each reference year can be refined specifying the kind of transport and accident causes. The figure 7 shows example of such a refinement of accident statistics for Automatic Half Barrier (AHB) level crossing from the year 2004.

Figure 7 Refinement of entered level crossing accident statistics (source: RSSB, UK)

Level crossing accident data analysis

The analysis tools of the knowledge management system allow to compare:

• the level crossing risk of different countries

• the risk of different level crossing types (according to the classification of ERA)

• the trends of level crossing accidents (in different countries and for different level crossing types)

• the causes of level crossing accidents

In the following sections examples of the analysis results will be given. These have been obtained on the base of the collected data from SELCAT partners. The actual data on level crossing accidents are representative for 7 European and 4 non-European countries. The data from some other countries are either not completed or they have not the required degree of detail, necessary for the stated aims of the analysis. As an example the collected data related to the year 2005 have been applied.

Level crossing risk of different countries

The first chart on figure 8 shows the total number of accidents, fatalities and injuries in particular EU countries and in EU countries together with non-EU countries which has been collected in the SELCAT project. It is clearly visible that due to the substantially different size and population of particular countries, their railway and road networks length and density, the number of accidents on the level crossings is so different that the large number of cases in some countries makes the data of some other countries invisible.



PROCEEDINGS 26/208

Figure 8: Absolute number of accidents in EU and EU + non-EU countries in 2005

As it can be seen the absolute numbers of accident cannot be used directly to compare of the level crossing risk. For this purpose a set of simple and complex normalisation (scaling) factors has been implemented in the knowledge management system. The figure 9 shows the corresponding menu allowing to set the desired normalisation factor for accident data analysis and visualisation. A multiple choice of simple normalization factors is possible.

0

500

1000

1500

2000

2500

Acc Fat Inj Acc Fat Inj

EU EU + non EU

Russian Federation

Morocco

Japan

China

United Kingdom

Slovakia

Poland

Hungary

Germany

Finland

Bulgaria

Accidents



PROCEEDINGS 27/208

Figure 9: Example of parameterization using the interface for visualization of accident statistics



PROCEEDINGS 28/208

The Fig. 10 shows application of the mostly used normalisation factor which is the level crossing population of the particular type. Using this kind of visualisation diagram considering the chosen scaling factor the countries with the highest level crossing risk can be identified. Similar visualization can be obtained for each basic level crossing type.

Figure 10: Comparison of accident statistics normalised by LC type population (nfs1)? and related to the countries in 2005

Risk of different Level crossing types

A further task of the accident statistics analysis carried out in the SELCAT project was to compare the operational risk between different level crossing types. According to the mentioned 7 basic level crossing types defined by ERA have been taken [1], only 5 of these types could be identified, when analysing all the collected national level crossing types of countries involved in SELCAT. Figure 11 shows the type related comparison of the accident statistics normalized by its individual level crossing type population.

0

0,01

0,02

0,03

0,04

0,05

0,06

0,07

0,08

0,09

Acc Fat Inj Acc Fat Inj Acc Fat Inj Acc Fat Inj Acc Fat Inj Acc Fat Inj Acc Fat Inj Acc Fat Inj Acc Fat Inj Acc Fat Inj Acc Fat Inj

Bulgaria Finland Germany Hungary Poland Slovakia United

Kingdom

China Japan Morocco Russian

Federation

B

A 2.3

A 2.1

A 1.3

A 1.2

Accidents / Number of LCs



PROCEEDINGS 29/208

Figure 11: Comparison of absolute accident statistics related to the countries

Comparison of trends of level crossing accidents

The SELCAT knowledge management system allows the visualisation of accident statistics trends using any simple or composed normalisation factor. These trends can be related to a particular country or to a particular level crossings base type (e.g. according to the ERA definition). The figure 12 shows the comparison of LC accident statistics trends on example of passive level crossings normalised by the number of million train kilometres. The change of the population during the visualised time period hasn’t been taken into account.

Accidents / Number of LCs



PROCEEDINGS 30/208

Figure 12: Accident statistics of passive level crossings normalized by the volume of train traffic (in millions train

kilometers) Comparison of causes of level crossing accidents

As the experience has shown most difficulties were related to the collection of data concerning the causes of the level crossing accidents. This information is mostly available only in accident reports which evaluation had to be carried out mostly manually. Despite this there could be collected data concerning the accident causes from 7 European and 4 non European countries.

Due to the wide range of causes and different categories and detail levels used in national databases it was necessary to introduce a harmonized way for classification. For this reason the SELCAT knowledge management system offers 5 categories of causes:

• Human causes road sides – including all intentional and non intentional road vehicle driver errors e.g. “zig-zaging”, warning light violation due to inattention, sun shine,

• Human causes rail sides – representing failures of railway staff e.g. gate keeper, train driver etc. • Technical causes road side – covering accidents caused by technical deficiencies of the road vehicles

e.g. grounding, deficiencies of brakes, wheels etc. • Technical causes rail side – including accidents caused by level crossing equipment failure (lights,

barriers etc.), • Other causes – covering all accident due to causes non fitting in any previous category e.g. accidents

caused by animals etc. The figure 13 shows example of visualisation of statistics on level crossing accident causes. More detailed analysis can be carried out using the level crossing type specific visualisation of the collected data (figure 14).



PROCEEDINGS 31/208

19 482

882

23

45

211

1152

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

EU non EU

Accidents human causes roadside

Accidents human causes railside

Accidents technical causes roadside

Accidents technical causes railside

Accidents others

Figure 13: Proportion of causes for level crossing accidents

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Accidents human causes roadside

Accidents human causes railside

Accidents technical causes roadside

Accidents technical causes railside

Accidents others

Figure 14: Proportion of causes for level crossing accidents on different level crossing types

As it can be seen the human causes on the side of the road traffic have the strong dominance for level crossing accidents. Their more detailed analysis brings important hints for development of effective measures for increase of level crossing safety. However any more detailed conclusions can be obtained only on the base of analysis of level crossing accidents reports. From this reason one of the future planned development of the Level Crossing Knowledge Management system is the implementation of an advanced user interface allowing directly to upload a accident reports of single accident. Such a kind of database would represent an alternative source for the presented visualization modes of the collected data.



PROCEEDINGS 32/208

Conclusions In the management of railway safety learning lessons from accidents is the key to understanding of risk, as well as to regulation and implementation of new safety initiatives.

The described functionality of the SELCAT Knowledge Management System has been designed according to the requirements of the safety performance monitoring necessary for practical application risk based safety philosophy declared in the Safety Directive 2004/49. It allows collecting national level crossing accident statistics coming from different statistic sources related mostly to the typical national level crossing types and a fair comparison free of national and specific circumstances. The common level crossing functional model is the building base for harmonized information structure making possible comparisons and global evaluations. These are crucial for determination of Commons Safety Targets which are to be defined and to be used for risk assessment of future European railway control and protection systems according to the mentioned Safety Directive.

A level crossing is a suitable example of a railway protection system for application of new safety regulations. The statistics shows that it possesses significant part of the railway risk. Its effective reduction will bring direct benefits in saved human lives and human health. Further completion of the statistic database of SELCAT KMS will allow more precise evaluations and comparisons. These results represent input for further work in the area of level crossing risk management.

The contemporary SELCAT KMS is designed for collection of national statistics evaluated on the base of accident reports (e.g. by railway infrastructure managers, national safety authorities etc.). These statistics often do not contain data of necessary level of detail about the occurred accidents, e.g. the causes, the severity, operational conditions or technical composition of particular protection system etc. However these play a significant role at identification and quantification of relevant hazards leading to the operational risk. Therefore the design concept of SELCAT KMS considers the future possibility of direct collecting of accident reports based on a common level crossing accident reporting protocol. Its specification is one of the outstanding aims of the SELCAT project.

Acknowledgment

The authors thank to the partners of the project SELCAT project for providing of the level crossing data used as exemplar results of safety performance analysis. References

[1] http://europa.eu.int/comm/transport/road/roadsafety/roadinfra/levelcrossings/index_en.htm

[2] Safety at Level Crossings, EC DG TREN, High Level Group Road Safety, 2003

[3] EN 50126: Railway applications - The specification and demonstration of reliability, availability, maintainability and safety (RAMS). CENELEC, Brussels, 1998

[4] Safety Directive 2004/49/EC of the European Parliament and of the council of 29 April 2004

[5] M. Meyer zu Hörste: Methodische Analyse und generische Modellierung von Eisenbahnleit- und -sicherungssysteme. Dissertation, Technische Universität Braunschweig, 2003.

[6] J. Drewes, R. Slovak, L. Tordai, E. Schnieder: Formal Structuring for Development of Level Crossing Ontology. FORMS/FORMAT 2007 – Proceedings of Formal Methods for Automation and Safety in Railway and Automotive Systems (G. Tarnai and E. Schnieder Eds.), Braunschweig, 2007, pp 355-360

[7] E. Schnieder: Control for traffic safety - safety of traffic control. In: Tsugawa, S.; Aoki, M., Hrsg.: CTS 2003 - Preprints, S. 1-13, Tokyo, Japan, August 2003. 10th IFAC Symposium on Control in Transportation System/Tokyo, 2003.

[8] A. Pira: Monitoring of Safety Performance activity “DEFINITIONS OF COMMON SAFETY INDICATORS on level crossings” (Draft), European Railway Agency, 2007



PROCEEDINGS 33/208


Development, Implementation and Use of the All Leve l Crossing Risk Model

Author(s): Alan Symons, Brian Tomlinson

Job Title: System Risk Control Specialist

Company: Network Rail Infrastructure Ltd


Resume of Speaker

Alan is currently the System Risk Control Specialist within Network Rail. He is responsible for supporting the development of corporate strategies for controlling key system safety risks including level crossing, SPAD, irregular working and route crime risks. Most recently this has been in the development of the All Level Crossing Risk Model. Alan has worked on the railways since 1981 as an operator and risk specialist providing risk expertise. He has previously been responsible for the development and implementation of the Signal assessment Toolkit for the assessment of SPAD risk and development of the original Safety Risk Model used by the industry.

Abstract

The presentation will cover the advances that have been made in the management of level crossing safety following the development, launch, ongoing population and use of the All Level Crossing Risk Model (ALCRM) in Great Britain. The ALCRM is now a key part of Networks Rail’s processes for the assessment, quantification and management of risk at levels crossings throughout the network. It is used as a tool to assist in the prioritisation of investment to reduce risk so far as is reasonably practicable.

Background There are around 7000 level crossings in active use on Network Rail managed infrastructure. Of these approximately 1500 are on public vehicular roads and the remainder are where public footpaths, bridleways and private roads/tracks cross the railway. Some private vehicular crossings have public footpath or bridleway rights. The layout, configuration and use of level crossings vary from location to location, so each one is essentially unique. To minimise the risk of trains striking crossing users the following features may also be present: • barriers or gates to physically prevent vehicle or pedestrian users from crossing the railway • coloured lights to provide a visual indication to the user of whether, or not, it is safe to cross; these may

also be combined with an audible alarm



PROCEEDINGS 34/208

• telephones for users to request permission from the signaller to cross • gates or stiles to highlight to the user where the boundary with the railway begins and end. These can

also prevent inadvertent trespass of children or animals onto the crossing or, in the case of locked gates, unauthorized use

• signage to explain the safe method of using the crossing or to bring the user’s attention to specific dangers

• railway signals that can be set to stop trains on the approach to crossings which are open to crossing users

• railway signs that signify trains to stop on the approach to crossings or to remind the driver to sound the horn

Exactly which of these crossing features need to be provided have for many years been specified by legislative requirements and industry standards, supplemented by Her Majesty’s Railway Inspectorate (HMRI) guidance. The principal factors that influence the requirements are maximum train speed, train frequency, crossing user frequency and whether it is for public or private use. Level Crossing Risk Level crossings are safe if used correctly. Over 90% of risk in the previous five years has resulted from user misuse in the form of error or abuse – the remainder being due to other causes such as equipment failure, reduced visibility or railway operator error. Typical examples of user error include incorrect knowledge of operation, misjudging the time it takes the train to reach the crossing or making incorrect assumptions regarding who has priority of use, direction of travel or the presence of a second train approaching - usually from the opposite direction. Typical examples of user abuse include users driving around half-barriers, users crossing when the crossing lights are red, users not requesting the signaller’s authority to cross (where required) and leaving gates open after use. On average there are seven pedestrian and two to three vehicle occupant fatalities per year (excluding suicides). Accidents involving injury to persons on the train are rare. However, the ever present risk was highlighted in 2004 when a train derailed following a collision with a car that had been deliberately parked on the automatic half-barrier level crossing at Ufton Nervet, Berkshire resulting in seven fatalities; the vehicle occupant, five passengers and the train driver. Other risks that arise at level crossings include user slips/trips/falls (including cyclists), trespass along the railway line itself, equipment damage due to vandalism, electric shock from overhead wires and vehicle collisions with barriers, pedestrians or other vehicles. The most effective way of reducing level crossing risk is to eliminate the crossing completely. Whilst purely private level crossings can be closed by agreement with authorised users, closure of public level crossings is notoriously more difficult under the present law. In addition, closure of a public level crossing may result in a requirement to provide an alterative route either in the form of a bridge over the railway, an underpass beneath it or through provision of a diversionary route to a nearby existing bridge, underpass or level crossing. Provision of structures such as bridges or underpasses involves large capital investment. It can also take a long period of time before they are realised due to the need to obtain the necessary planning (and other) consents and the magnitude of the infrastructure works required. Additional land may also need to be purchased. Network Rail is subject to the requirement of the Health and Safety at Work Act etc 1974 to reduce risk ‘so far as is reasonably practicable’. In simple terms this means that the cost, time and effort required in providing a specific risk reduction measure needs to be commensurate with the safety benefit that will be obtained as a result of its implementation. In the majority of cases the risk associated with individual level crossing use is insufficient to make a clear case for its closure and/or diversion. It is therefore necessary to understand other benefits that can be factored in, for example reduced operational or maintenance costs, avoidance of forthcoming renewal costs, improved operating performance or funding obtained from other parties involved such as the Highway Agency, local councils or private housing developers. Management judgement also forms a key part of the decision process



PROCEEDINGS 35/208

when qualitatively the risk warrants something to be done but the case for closure and/or diversion is not necessarily clear cut. If it is not practicable to close and/or divert the crossing then it may still be possible to reduce risk through the provision of improved safety features where it is considered reasonably practicable. In contrast provision of new level crossings would introduce additional risk and therefore would be permitted only in exceptional circumstances. Risk Management Strategy Network Rail’s overall strategy for managing level crossing risk is based upon a principle known as the four ‘E’s: • Education; educating crossing users on how to use level crossings correctly and highlighting the dangers

of misuse • Enforcement; taking appropriate action to assist the police in identifying those who deliberately endanger

others through their actions at level crossings with a view to securing their prosecution • Engineering; requiring that level crossings are regularly inspected and correctly maintained. Additionally,

where it is reasonably practicable to do so, enhancing crossing safety through means such as closure/diversion or provision of additional safety features/equipment (e.g. addition of telephones or lights, conversion from half-barriers to full-barriers)

• Enablement; developing appropriate techniques, processes, models and relationships/partnerships to improve the management of level crossing risk (e.g. Road Rail Partnership Groups)

With such a large population of crossings it is essential that resources are targeted towards managing risk at those crossings which present the greatest risk, and that the risk reduction measures proposed are commensurate with the safety benefit obtained. This is where the ALCRM comes in.

Development of the ALCRM The All Level Crossing Risk Model (ALCRM) was developed as a means to understand the relative risk that each crossing on the network presents and to evaluate options to ensure risk is reduced so far as is reasonably practicable. Prior to the existence of the ALCRM there had been a number of separate methods of assessing level crossing risk. The most advanced of the time was spreadsheet-based though it was used only for assessing the risk at automatic level crossings. The remainder of the level crossings on the network were risk assessed using either a simple form-based scoring system or qualitatively based upon observations made during their cyclic inspections. In essence, the approach was fragmented and did not provide a holistic picture of level crossing risk. In addition records were held locally and were not able to be shared consistently throughout the company. The key objective of the ALCRM was to provide a single means of understanding the relative risk presented by each of the 7000 level crossings on the network. It was also to provide a standard means for collecting, storing and recalling data on individual level crossing features and to permit the consideration of different risk reduction solutions such that the most appropriate could be selected from both a safety and business perspective. Additionally the objective was to develop a model that incorporated the best knowledge on level crossing risk that was available to date such that, by being one of the most sophisticated models in use internationally, it would be regarded as a world class tool for managing level crossing risk. A dedicated team effort has enabled the vision of the ALCRM to be realised. The Rail Safety & Standards Board (RSSB), working in collaboration with Network Rail, took the lead role for the research part of the project and enrolled the expertise of Arthur D Little to devise the complex and extensive algorithms/formulae which



PROCEEDINGS 36/208

form the core workings of the model. Network Rail worked with both parties sharing their operating experience and knowledge of level crossings to ensure the model closely reflected reality. Network Rail also took the lead role in the translation of the model algorithms into a software based tool on a central Oracle based database system, involving specialist information management consultants Strategic Thought Limited. Network Rail also provided the training and support to its own area based teams who are now responsible for data collection, input to the model and use of it as a decision support tool to manage the level crossing risk results it provides. Migration from the existing risk assessment processes to the ALCRM also required a wholesale revision of Network Rail’s procedures and processes regarding level crossing risk. The ‘production’ version model was successfully launched on Network Rail’s information management system on 12 January 2007. How does it work ? The ALCRM is a database software application. It is hosted on the company mainframe system and access can be gained by Network Rail staff, who have been trained and given the required permission, from their desktop computers. There are three levels of permission: user (can add/view/edit records), viewer (can view records only) and administrator (can add/view/edit/delete records, set-up/amend/delete user accounts and amend configuration/calibration data). The ALCRM is able to calculate the level of risk each crossing presents, using a series of complex algorithms, in response to data relating to the crossing input by the user. The ALCRM holds a ‘live’ risk assessment for each level crossing that is entered (i.e. the current risk at that level crossing). In addition, users are able to modify the input data to understand how the risk would be altered through the identification of potential risk mitigation options (e.g. close crossing, add additional protection measures, convert to a different type). It is also able to calculate the change in risk should a new timetable be proposed or in the event of a forthcoming development in the area that could change the number/type/pattern of crossing users. Any other changes that could affect the level crossing input data can be modeled in this way to determine how the risk is affected. Should any of the risk mitigation options, timetable changes or developments be taken forward the ALCRM will update the live risk assessment when the user enters that the scheme has been implemented. A history of any previous risk assessments is held until such a time as it is necessary for the data to be archived. By comparing (a) the current risk with (b) the risk of a potential risk mitigation option the ALCRM is able to calculate the monetary safety benefit or loss that would be obtained. This is achieved using the Value for Preventing a Fatality (VPF) for the rail industry which is published annually by RSSB. The ALCRM has an in-built Cost Benefit Analysis (CBA) tool that can determine the cost/benefit ratio of the proposed risk mitigation option. In addition to safety benefit, the CBA tool is also able to take into account the capital investment and ongoing operational or maintenance benefits/cost, which are discounted, over the life cycle of the proposed scheme. Data Collection and Input Network Rail procedures require that each level crossing is risk assessed every three years as a minimum. In addition annual censuses are undertaken for each public road crossing. There are a number of other criteria that can trigger the need for an assessment during the course of the three yearly cycle. These are: • significant timetable change • substantial (10%) increase in road traffic • following instance of misuse, near miss or accident • at the evaluation stage for new level crossings, proposed renewal or alteration to protection type • planning proposals consultations that indicate substantial change in road traffic volumes, patterns or

speeds • significant change in environment on approach to crossing • following formal expression of concern by Network Rail, Train Operator or Highway Authority



PROCEEDINGS 37/208

Data for the risk assessment process is collected by means of a site visit to the level crossing. This is usually undertaken by a Network Rail Mobile Operations Manager (MOM). Around 200 items of data are collected relating to each level crossing. The principle items include: • crossing name/type/geographic location • frequency/speed/type/length of trains • frequency/type/familiarity of crossing users; vehicular, pedestrians or animals • physical aspects relating to the crossing particularly those that can affect its safety such as sighting

distances, type of surface, number of lines or direction in relation to the sun • physical features in proximity to the crossing that could affect the consequences of train derailment e.g.

tunnel, viaduct, station, points • factors that indicate a higher likelihood of misuse or violation (e.g. known problems, recent accident). The data is recorded onto a paper form that, when complete, is returned to the local Area office. A trained and competent user, usually the Level Crossing Risk Control Coordinator (LCRCC), then manually keys the data on the form into the ALCRM. A project is currently underway to look at whether the data can be captured using a handheld computer so that there is no need for it to be manually input into the system. Once all the data has been input a second trained and competent user must then sign-off the data to confirm it has been sense-checked for potential errors. Once the data has been signed-off the ALCRM then calculates the risk. This is known as the ‘live’ assessment. ALCRM Output For each level crossing the ALCRM is able to calculate: • the Collective Risk to the exposed populations (i.e. the user and anyone else involved as a result of any

accident such as passengers/staff on trains); this is expressed in Fatalities & Weighted Injuries (FWI) per annum (where 1 fatality = 10 major injuries = 200 minor injuries)

• the Individual Risk to the user; this expressed as a probability of a fatality per year based on an average user traversing the crossing 500 times in one year (e.g. 1 in 240,000)

• the apportionment of both collective and individual risk across the types of users and also the risk to passengers and staff on the train

The ALCRM output also includes: • a list of the key factors that influence the risk score, known as key risk drivers (e.g. poor sighting time,

large number of users, use by Heavy Goods Vehicles) • a risk categorisation to simplify understanding based on a scale of 1 (high) to 13 (low) for Collective Risk

and A (high) to M (low) for Individual Risk; e.g. a crossing with a risk score of B2 is categorised as higher risk whilst a crossing with a risk score of L11 falls into the lower risk bracket.

‘Optioneering’ All crossings will ultimately undergo an ‘optioneering’ exercise to identify what potential risk migration options exist. To make sure that effort is initially concentrated towards those crossings that present the greater risk the LCRCCs are instructed to focus initially on those with a Collective Risk ranking of 1, 2 or 3 (higher risk). Then, to focus on those with a Collective Risk ranking of 4 or 5 or an Individual Risk ranking of A, B or C (medium risk). The ‘optioneering’ exercise first involves a site visit by the LCRCC, accompanied by other parties as necessary (e.g. engineer, authorised users, local council representative, Highways Agency representative). The purpose



PROCEEDINGS 38/208

of the visit is to determine the potential risk mitigations options (if any) that are available e.g. is closure feasible? Installation of telephones, improved signage. A web-based application, known as the Level Crossing Risk Management Toolkit is available as a prompt in the identification of potential mitigation options. On returning to the office the LCRCC inputs each of the identified options through copying the ‘live’ assessment and modifying the crossing input data. The ALCRM is then able to calculate the level crossing risk if the option were to be implemented. The CBA tool can then assist in the decision making process as to whether, or not, it is reasonably practicable to implement. If a reasonably practicable mitigation option is identified the LCRCC is able to recommend to the Operations Risk Advisor (ORA) that it is taken forward. If the ORA is in agreement then a formal business case is prepared and investment sought through the relevant Route Director or Network Rail headquarters as necessary. Once the funding has been obtained the selected option is marked as ‘approved’ within the ALCRM. When the works have been implemented the ALCRM replaces the ‘live’ assessment with the risk assessment for the selected mitigation option. Unless the crossing is closed then the three yearly risk assessment cycle continues. Progress to Date A target has been set to achieve 100% population of the ALCRM within three years of its launch (11th January 2010). Initially, population of the ALCRM was slow, because the MOMs first had to be trained on the data collection requirements. The users also had to become familiar with the new system and transfer over from their locally held assessment schedules/records to the central data repository. However, through weekly monitoring and reporting of progress it was soon possible to direct efforts to assist those in most need. Within six months the population rate of the ALCRM was over 100 crossings per week. By the end of 2007 all public road crossings and all station foot crossings had been entered into the ALCRM. At the time of writing (end-Apr 2008) over 70% of crossings have been entered; therefore it looks as if the initial target will be met, if not beaten. Whilst the initial focus of the monitoring was ‘population’, the emphasis has now moved to making sure that the crossings with a higher Collective Risk ranking have been ‘optioneered’. Around 15% of the total population of crossings entered to date have a Collective Risk ranking of 1, 2 or 3, and of these, at the time of writing, approximately 45% show evidence of optioneering having been undertaken. In some cases there are few or no mitigation options available. However, for some crossings, the ALCRM is proving to be a valuable tool. For example at Quay Ward No2 user-worked crossing three potential mitigation options were identified; (i) move gates (ii) move gates and install additional signage or (iii) close. The crossing had a Collective Risk ranking of 4 (medium risk) equating to 0.0017834 FWI per annum. The ALCRM demonstrated it would be cost effective to select mitigation option (ii) as it would provide a £33,000 safety benefit compared to the cost of £12,400 of moving the gates and providing additional signage (safety benefit : cost ratio of 2.66 : 1). Other Benefits The ALCRM is also proving to be of wider use in the management of level crossing risk. Two examples of this are as follows: • the ALCRM stores the Ordinance Survey co-ordinates for each crossing location and also identifies

whether user abuse is a known problem. Using data extracted from the ALCRM, and geospatial mapping software it was possible to produce a map showing the location of crossings that are abused and their risk ranking. The map has been made available to the team who manage the ‘Don’t Run The Risk’ level crossing public safety campaign to assist them in where best to target investment in television, radio and newspaper features/adverts.

• the ALCRM Collective Risk rankings have been provided to the national team responsible for managing the level crossing renewals programme such that they are able to understand those crossings that present the greater risk. Efforts can therefore be concentrated on identifying whether a closure solution



PROCEEDINGS 39/208

exists, sufficiently enough in advance, so that the costs of the forthcoming renewal of the crossing might be factored into the business case for closure.

Calibration and Enhancements The algorithms within the ALCRM are based upon a series of fault trees that are derived from data contained within the industry’s Safety Risk Model (SRM) which is managed by RSSB. The data contained in the SRM is based on historical accident/incident data although it is also predictive to some degree. As the data in SRM changes with time so must the ALCRM in order to remain aligned to it. Therefore the ALCRM contains a number of adjustable calibration constants/variables that can be amended by the system administrator. In risk terms, the initial calibration of the ALCRM is not too different to the SRM - in most cases it is slightly overestimating risk. A calibration programme is currently underway to bring them more closely into alignment. In addition, in response to user feedback, an independent review and other suggestions a series of enhancements are proposed that will improve the systems functionality and also it’s modeling capability. These are planned for the next two years. Summary Through use of the ALCRM, Network Rail is now able to understand the level crossings which present the greater risk, so efforts can be targeted to close/divert or improve these crossings where reasonably practicable. Having one model which is used across the whole company, enables the industry to manage level crossing risk at both a strategic national level as well as a local tactical level in the knowledge that consistent standards and processes are being applied throughout. …and Finally The ALCRM was awarded the Institute of Engineering Technology (IET) 2007 Award for the Advancement of Railway System Safety. Acknowledgements The authors would like to acknowledge the following: • John McMorrow and Michael Woods from RSSB, and Marcus Beard, John Barker and Jay Heavysides

from AD Little, for the development of the model algorithms • Warrick Dent, Gilbert Fraser and the Area based operations teams within Network Rail Operations &

Customers Services for their vast knowledge and experience of level crossings and for the wholesale revision of the company’s risk management processes

• Jon Alcock, Neil Mitchard and Phillip Wheating from Network Rail Information Management, and Ben Savage from Strategic Thought Limited, for translating the algorithms into a highly reliable database application

• Pete Stanton and Sam Pead from Network Rail, and Ben Gilmartin from RSSB, for their assistance in providing sound technical risk advice and in testing the system.



PROCEEDINGS 40/208

Session 2 - Risk Management Chairman: Michael Woods, RSSB, UK

Recent experience in Victoria, Australia

Author(s): Alan Ross

Job Title: Principal

Company: A&K Ross Associates Pty Ltd

Country: Australia

Resume of Speaker

Alan is currently a Principal of A&K Ross Associates Pty Ltd (AKRA). He began his working life as a front-line pilot with the Royal Air Force (UK), based in Germany at the height of the ‘Cold War’. Here he had his first experience of risk management - something called MAD – ‘Mutually Assured Destruction’. He later joined the production side of the offshore oil industry in the North Sea, during an expansion that continually pushed the ‘envelope’ of what was possible. He has 25 years experience in safety and risk management, having worked in the oil and gas industry and the mining, construction, petrochemical, aviation and rail industries in UK, Australia and Middle East. He was Public Transport Safety Regulator in Victoria from 1998-2000. He is a Chartered Member of the Institution of Occupational Safety & Health. Alan has over 50 clients in a range of industries. He is a self-proclaimed ‘student of disaster’, a futurologist and astronomer. Alan has extensive experience of level crossing risk assessment and design, and has recently facilitated 15 Design Risk Workshops, involving over 50 level crossings of various types, as part of the State Government of Victoria’s upgrade programme. Abstract The State Government of Victoria, in Australia, has initiated a greatly accelerated level crossing upgrade programme following a crash between a heavy road vehicle (semi-trailer) and a locomotive hauled passenger train in which 11 rail passengers died. The level crossing location was at Kerang, in northern Victoria, a single track lightly used line and a crossing that was protected by flashing lights and bells only. This was not an isolated event and the costs are large. This paper will provide a brief synopsis of the crash, as well as others at level crossings in Victoria in the last 18 months, all involving fatalities and/or heavy road vehicles. The options available to deal with this situation will be explored, including costs and practicalities. With a lateral approach some new ideas will be floated and discussed, as well as the underlying causes for many of the crashes, the solution to which may be beyond the control of any rail organization. 1. The scope of the subject

Accidents are rarely accidental

He heard it, but he heeded not – his eyes were in his heart, and that was far away. Byron, Childe Harold, cxli



PROCEEDINGS 41/208

In May 2006 a truck crashed into a freight train near Lismore in Victoria. The level crossing had passive warning signage only. At the time visibility was poor due to morning fog. The train was hauled by three locomotives, was 1356 metres long, weighed 4382 tonnes and was travelling at 112 km/h at the time of the collision. The truck and trailer weighed 48 tonnes and its estimated speed at impact was between 53 and 78 km/h2. The truck collided with the side of the second locomotive. The force of the impact was sufficient to derail this locomotive and the third locomotive. This led to the catastrophic derailment of 41 of the 64 wagons, which were compressed into an area 128m long by 45m wide and 12m high. The truck driver was killed. The recovery operation took several days, causing severe disruption to interstate rail traffic. Added to the losses from the crash and other costs the total cost of the crash was later estimated at between A$30 and A$40 million. The truck company had $10 million of insurance coverage and there were a number of claimants, so clearly full recovery was not possible, leaving the claimants to seek the remaining costs through their own insurance. This situation is not uncommon and obviously can/does lead to higher premiums for the rail industry. The investigation report by the Australian Transport Safety Bureau (ATSB) did not recommend the closure of the level crossing, despite the fact that there is another crossing some 300m away and adding a spur road would be straightforward. The exact number of level crossings in Victoria is probably not accurately known, because, in addition to the public crossing points, there are many, many private crossings, sometimes known as ‘occupation crossings’ plus a number of ‘informal’ crossings. At present in Victoria there are officially around 2250 public level crossings of one kind or another. This represents approximately a quarter of the national total.

2 Australian Transport Safety Bureau Investigation Report



PROCEEDINGS 42/208

These level crossings are located on quiet country roads and in areas of high-density metropolitan traffic. On the quiet roads the road user is generally in a low state of attentiveness, in the latter the road user could very well be overstressed. In both these cases, the performance of the road driver is potentially going to be sub-optimal . How else do you explain why a truck drives into the third car of a passenger train, in broad daylight, good visibility and with the flashing lights on the level crossing operating? Does this situation lead to a possible need for additional and/or different types of ‘protection’ at crossings? If so, what? The State Government of Victoria, in Australia, has initiated a greatly accelerated level crossing upgrade programme following the crash between a heavy road vehicle (semi-trailer) and a locomotive hauled passenger train in which 11 rail passengers died. The level crossing location was at Kerang, in northern Victoria, a single track lightly used line and a crossing that was protected by flashing lights and bells only. This paper will provide a brief synopsis of the Kerang crash, as well as five others at level crossings in Victoria in the last 2 years or so, involving fatalities and/or heavy road vehicles.

2. The issues at stake

In flying I have learned that carelessness and overconfidence are usually far more dangerous than deliberately

accepted risks.

— Wilbur Wright in a letter to his father, September 1900. Victoria has a large number of level crossings with only passive ‘protection’ or flashing lights/bells protection. Many of the latter are on country highways that are used increasingly by heavy road traffic and frequently have a road speed limit of 100km/h and a rail line speed of up to 115km/h. The prevention of further serious accidents such as the ones mentioned has become a major social, political and economic issue. The ‘outrage’ factor often seems to overrule the ‘idiot’ factor. The ‘menu’ of options available for level crossing protection has not increased significantly in recent years, especially where it is not economically feasible to install CCTV or such other higher technology devices. Grade separation is presently precluded in all but a handful of crossings in Victoria. There is no strategic plan for any grade separation, even in the metropolitan area. In the meantime road traffic is generally on the increase, often with a mix of vehicles that is increasingly heavier than in the past, and likewise rail traffic is often increasing at the same time. Negative or illegal road user behaviour is a factor in many of the accidents and this is demonstrably on the increase too. Consequently anything that requires a road user to be more patient and pay more attention is a problem. This is just what is required at a typical level crossing in Victoria but many road users are either not willing or not able to be more patient or pay more attention, or both. So the Nanny State steps in again and at the Australasian Railways Association Rail Safe Conference in February, the Victorian Minister for Transport announced that the road speed limit at level crossings would be reduced to 80km/h. At this point it is not clear if this is to be universal or selective, for example on highways. Assuming that grade separation is generally not considered a viable option (even if we disagree with that), what is the current menu of options available in Victoria to reduce level crossing risk?

• Passive signage & road markings • Flashing lights and bells (including use of brighter LED lights) • Boom barriers • Active signage • Rumble strips • Improved sight lines (for example by removal of vegetation)



PROCEEDINGS 43/208

3. The landscape before and post Kerang

You are only as good as your last disaster

On 5th June 2007, at Kerang in northern Victoria, a semi-trailer (articulated road truck) hit a 3-car locomotive hauled train. As a result 11 of the train passengers were fatally injured. The crash occurred at 1330 hrs and the weather was fine with good visibility. The level crossing was ‘protected’ by flashing lights, bells and passive warning signage. This was not an isolated event but what is often referred to as the ‘outrage factor’ kicked in and produced a significant response from the Government. Whether or not this response turns out to be a measured and optimal response in terms of the money being spent remains to be seen. As an aside, the truck driver, who survived, refused to be interviewed by the Investigator and was subsequently charged by Victoria Police with culpable driving. This paper will focus on what is being done in Victoria to reduce the risk, including some new ideas that are considered economically feasible. The paper will also focus on the various ways in which risk at level crossings is assessed, including relatively sophisticated assessment models that do not always seem to target the crossings at which accidents occur. Ways of enhancing the risk assessment process will be proposed, based clearly on a cost-benefit approach, but not necessarily the ‘ALARP’ approach that has become the norm in many parts of the world. By looking at a range of level crossing locations and trying to understand the variables it is hoped to provide the basis of a fresh approach to the problem. Before Kerang there existed in Victoria a level crossing upgrade programme with funding of around A$3m. Operating on established but possibly outdated criteria, crossings were upgraded on a priority basis, resulting in some 12-13 crossings being upgraded per year, either from passive to flashing lights, or from flashing lights to boom barriers. A survey of level crossings using a new tool called Australian Level Crossing Assessment Model (ALCAM) has been introduced. ALCAM assessed the Kerang crossing as ‘needing no attention’, prior to the crash in June of 2007, raising questions as to the efficacy of ALCAM. Post Kerang, with the ‘outrage factor’ kicked in, a range of measures was proposed as well as a greatly accelerated upgrade programme, based on ALCAM assessments. In the financial year to June 2008, 45 crossings will have been upgraded. Other measures applied include:

• Installation of ‘rumble strips’ at 200 country crossings (this is largely complete) • Installation of ‘Active Advanced Warning Signs’ on all ‘highway’ crossings – a total of 53, of which 5

have been installed. • Trial of CCTV cameras at two locations • Increased penalties for violations • Awareness programmes

As well as the announced reduction to 80km/h at crossings that previously had a 100km/h road speed limit. It is early days but indications are that this somewhat ‘knee-jerk’ response has not turned things around Since Kerang, 12 months ago, in Victoria there have been 8 more fatal crashes at crossings and several non-fatal crashes. One of the fatal crashes was at a crossing that had recently been fitted with rumble strips, causing the effectiveness of that measure to be questioned already. Interestingly, rumble strips were also installed at Kerang and there were reports of truck drivers driving round them, on the wrong side of the road! During recent times there has also been a sharp up-turn in the number of reported near misses at crossings. Despite these efforts the vast majority of crossings in Victoria remain ‘protected’ purely by passive signage and that situation is not going to change dramatically in the next 10-20 years. Some say that passive signage does not represent ‘protection’ as it gives no indication of a train approaching.



PROCEEDINGS 44/208

4. Other crashes briefly

4.1 April 2006 – Trawalla – 2 dead on the train – no report has been made available to the public 4.2 25 May 2006 – Lismore – truck driver dead – see ATSB Report 4.3 November 2006 – Cressy/Wingeel – Barpinba – Poorneet LX – truck driver dead – see ATSB Report 4.4 August 2007 – Somerville – truck driver dead 4.5 January 2008 - Mildura – no fatalities or injuries. 25,000 litres of Chardonnay wine lost. During this same period there have been numerous other fatalities at level crossings involving cars and pedestrians.

5. Potential for improvement?

Man is a battlefield, a dark cellar in which a well-bred spinster is locked in combat with a sex-crazed monkey, their struggle being refereed by a nervous bank clerk

Civilised man – Bannister, D. Some would also say that many, if not all, level crossings are simply accidents waiting to happen, even if they do have ‘protection’. If the current menu of improvement options is not adequate, what other options should be considered? Critics of the current programme would say that many of the measures have not been scientifically investigated – to determine their effectiveness. There is currently no formal research body in Australia looking at level crossing protection. However, there is certainly activity going on to consider other options, mostly on a speculative commercial basis. A body called Intelligent Transport Systems Australia, in conjunction with the Australasian Railways Association, recently ran a full day session looking at a range of options to improve level crossing protection. A brief description of some of them is given below: 5.1 GPS based train location system with local warning to road users via a radio beacon that triggers in vehicle radio, and locally based Global Frequency Warning that overrides active radio frequencies with a warning message of train approaching 5.2 Intelligent road studs that light up in conjunction with the other protection equipment 5.3 Intelligent CCTV – using existing CCTV installations more effectively and automatically to monitor and

warn of dangerous action + a CCTV automated enforcement system 5.4 Road user operated crossing gate (normally in the closed position) – for crossings that seeing little use,

with operation of the barrier varying depending on the traffic level, but returning to closed under gravity(Custom Traffic)

5.5 Active rumble strips – that operate only when a train is actually approaching 5.6 Improved warning lights that better gain the attention – strobe lights have been used elsewhere,

apparently to good effect. Consideration could also be given to installing them on trains. 5.7 Warning systems integrated with existing road vehicle technology such as satellite navigation systems

– linked to an accurate database of hotspots etc. 5.8 Use of road style stop-lights at level crossings – which are said to be obeyed more than rail style lights. 5.9 Use of an automated retractable barrier that is raised as part of the level crossing protection and

designed to physically stop a road vehicle (with damage!) Most of the above would be rejected on cost grounds, at least in the short term, for country crossings in Victoria. The way forward would seem to be a research based trial of some of the more promising and cost effective ideas. Many of these ideas may be already in use or under trial elsewhere. The role of this conference might be to provide some coordination of all such efforts such that we are not all continuously re-inventing the wheel! The details of Kerang and the other crashes are interesting and disturbing, as also is what the informed observer sees whilst travelling the roads of Victoria during the course of visits to crossings in connection with proposed upgrades, namely the author. Clearly this is not really a technological problem, not really a



PROCEEDINGS 45/208

behavioural problem, but more of a lifestyle problem. Everybody is in so much of a permanent rush, whilst at the same time suffering from information overload, that logical thought patterns are often suppressed. How many distractions are in the modern truck? Navigation system, stereo music/radio system, CB Radio, hands free cell-phone etc? What can be done about this? Unfortunately we face an impossible barrier here – lifestyle, technology and other developments have seriously outpaced human evolution! The human being has not fundamentally changed, either physiologically or psychologically, in 10,000 years. The amazing thing is that this primitive creature can drive a car or fly an aircraft, for most of the time, quite safely! Automation may be the answer – in commercial aviation many things are automated to improve safety. In Victoria, at a typical level crossing very little is automated, apart from the active protection that is installed at the relatively small proportion of crossings. In the abstract I suggested that nothing revolutionary will be proposed, rather a more lateral approach to using what is currently available, easily installed and maintained and aimed at maximum impact on the risk profile. I would like to suggest the following, which some may consider revolutionary:

• Graphical and hard-hitting advertisements that show quite clearly what can happen when road vehicles collide with trains. Before doing this there must be appropriate research to demonstrate that the techniques used will work, in getting the message across and achieving a positive change. The BBC programme ‘Top Gear’ recently showed a controlled crash in great detail – it would stand usage as a general safety promotion. Critics would argue that public safety awareness programmes do not work – if that is the case it is because they are wrongly focussed and were not properly evaluated for effectiveness.

• The truck driver at Kerang refused to be interviewed by the Rail Investigator, who had no powers to compel him. The investigating body must have such powers. They key to any solution to the ongoing carnage is to better understand why people do what they do. Maybe it is not simply stupidity, as is often assumed, for example the ‘Leibowitz Phenomenon’ has been suggested as a reason in at least some crashes – relative mis-perceptions of approach speeds of larger objects.

• Remove ‘Stop’ signs at level crossings (or at least most of them) – since there is something like 80% violation of these signs they in fact induce a view that such signs are not mandatory, even at road intersections. Additionally, in the case of modern large trucks, such as B-Doubles, if they stop it then takes longer for them to clear the crossing.

• Legislate a nation wide ban on any new at grade road/rail crossings. In Victoria this is nominally the case, requiring Ministerial approval, but there have been some new crossings despite this.

• Legislate the removal of crossings that are clearly completely or partially redundant – establish a list of such crossings based on practical criteria. There must be many crossings that are arguably redundant (perhaps with some change to the adjacent roads). Excuses as to why this is not practical have to viewed with a jaundiced eye.

• Legislate a nation wide programme of grade separation for exiting crossings with a stated target for total elimination (be that even 100 or 200 years time!) Grade separation is really the only complete solution. If there is no strategic plan to achieve that, the problem is never going to go away.

• As stated earlier, run a programme of trials, possibly for all the sensible, if unrealistically costly, ideas on improved protection.

• Finally, in a train/car collision the train normally ‘wins’, but in a train/truck collision this is increasingly not the case, with heavier and heavier trucks, lighter and faster trains. In most cases of a train-truck collision it would appear that the truck is the ‘aggressor’, accepting that the train cannot stop and has the right of way. Therefore it is essential and unavoidable that the trucking industry be engaged in a constructive way to see what can be done about this.



PROCEEDINGS 46/208

6. Added value to managing road safety at the road/ rail interface

‘They’re funny things, accidents. You never have them till you’re having them.’

Eeyore, ‘The House at Pooh Corner’

Loss of life is always a tragedy, but in addition it has to be acknowledged that the overall individual costs of some of the recent level crossing crashes in Victoria run into tens of millions of dollars. In the case of Lismore, the truck driver died but the total cost of the crash has been estimated at A$30 million. When there are several occurring each year this is obviously a totally unacceptable situation. However, expensive attempts at a rapid fix are not likely to eliminate the risk, so it is important that all parties work together proactively, to produce the required improvement. In Victoria a 50% reduction of accident/incident rate would more than pay for the current upgrade programme. Equally there has to be a more strategic view and a long-term programme of grade separation, at a rate that will dramatically reduce the overall risk. No such programme currently exists. By long-term I mean 50-100 years or more, with a risk based targeted programme that eliminates an agreed number of crossings every year. I am confident that an effective cost benefit can be shown for such a programme. Grade separation in metropolitan and other busy locations provides relief of road traffic congestion, means that less fuel is consumed by road users, they arrive at work sooner and they are less stressed, all of which clearly has both social and economic ongoing benefits.

7. Conclusions

Though the Life Force supplies us with its own purpose, it has no other brains to work with than those it has

painfully and imperfectly evolved in our heads

George Bernard Shaw – The Irrational Knot 1. From the point of view of economics, the most obvious conclusion is that level crossing accidents

generate costs far above and beyond what would generally be assumed the case. Although proportionally the percentage of lives lost on the roads due to level crossing crashes is perhaps 2% of the total, the total cost is likely to be much greater than 2%. Insurance is a growing factor in the management of costs and initiatives.

2. Existing methods of ‘protection’ are becoming less effective, because of heavier road vehicles and lighter, faster (passenger) trains. New technology must be found, tested and used. A co-ordinated research programme is needed that should include the insurance industry.

3. Grade separation is the only real answer and there must be a strategic plan to achieve that. 4. Whatever is done must fully recognise the limitations of the Mark 1 Human Being. We cannot hope to

reverse the increasing ‘rat-race’ that we are imposing upon our societies and people are going to remain increasingly impatient. They will not obey a sign if they don’t think they need to.

8. Contact details

Alan Ross, Principal A & K Ross Associates Pty Ltd, PO Box 283, Kangaroo Ground, VIC 3097, Australia Web: www.akra.com.au Email: [email protected]



PROCEEDINGS 47/208


Statistical analysis of risk

Author(s): Dr. Chris Elliott

Job Title: Director

Company: Pitchill Consulting Ltd

Country: United Kingdom Resume of Speaker

Chris Elliott is an engineer and lawyer, working with industry and government to help solve problems where regulation and technology conflict. Much of his work is with rail. He was an adviser to the UIC Safety Platform from its inception until the end of 2007, has worked extensively for RSSB in the UK and has been a member of several specialist teams, ranging from accident investigation to programme planning. He also has worked with rail authorities in other countries. He is a non-executive Director of the Office of Rail Regulation, the UK’s combined economic and safety regulator (and hence National Safety Authority).

Abstract For the first time there will be comparable data for railway risks in many countries. This is in part, at least for Europe, as a result of the Railway Safety Directive which requires data on which to base Common Safety Indicators and Common Safety Targets. This paper examines the data that is becoming available and its reliability, then uses it to identify statistical links between possible causes of risk and the consequences. The results do not prove a causal link but can identify the features of LX that are correlated with the level of risk and therefore provide guidance when seeking to improve management of the road/rail interface. It also comments on the challenge of defining Common Safety Indicators and Common Safety Targets. The principal conclusions are:

• data is available but there are omissions, not helped by inappropriate secrecy in some cases • the best predictors of LX accident rates are the behaviour of road users, suggesting that improving LX

safety is primarily an issue for roads authorities • indicators intended to compare the relative safety performance of countries must be carefully designed

to be robust and defensible, especially is used to inform legally binding targets.

Introduction

Accidents at level crossings, between trains and road vehicles, are perceived as a rail issue which must be addressed by actions by the railway company. In principle this is not unreasonable; LX accidents contribute a large fraction of the safety risk of railways. Conversely, LX accidents represent a much smaller fraction of the risk of road travel so the road authorities are reluctant to divert resources to reducing the risk.



PROCEEDINGS 48/208

There is little point in demanding action by the railways if the cause of the accident lies out of the railways’ control. It is hard to prove where the responsibility lies but it is possible to draw inferences from the statistics of those accidents. The Rail Safety Directive requires national authorities to submit data on the numbers of people killed or seriously injured at LX each year, and the number of LX. There is also a Common Safety Target for safety at LX. However, those statistics alone are not enough to indicate why LX accidents occur.

Approach

This paper seeks to find the factors that are correlated with LX accident rates. Although the existence of a correlation does not necessarily imply a causal link, it is not unreasonable to assume that there is at least a common root cause for the safety of LX and the factors which are correlated with it. This is analogous to the use in medical research of “risk factors”. For example, obesity is a risk factor for heart disease in that obese people are more likely to suffer heart disease. It may be that heart disease is caused by obesity; it may be that both heart disease and obesity have a common root cause. In either case, the existence of the correlation provides researchers with pointers to the aspects of diet, lifestyle, genetic makeup and other factors to explore. The approach taken here is to seek the combination of factors that is the best predictor of LX risk. The steps are: 1. identify a measure of risk, such as the number of people killed in LX accidents per year 2. identify a combination of risk factors that might be a good predictor (taking care that the combination has

itself scales properly; a good test is to imagine two identical countries being merged – the Safety Indicator should not change)

3. calculate Safety Indicators, defined as the measured level of risk in each country divided by a scaling factor which is a combination of factors for that country. Formally, the Safety Indicator for country m is referred to as SIm and is defined by SIm = Fm / Sm where:

- Fm the average number of LX fatalities per year in country m

- Sm the scaling parameter for country m

For example, one simple safety indicator is (a measure of risk) divided by the number of LX in the country (a scaling factor).

4. calculate the spread in values of SIm across all countries for which data is available. Two measures of spread are considered: standard deviation of the values and a weighted measure of deviation that gives greater significance to larger countries.

A low value of spread implies a good correlation and hence that the chosen SI is a good predictor of the level of safety risk.

Availability of data

There is data for the number of people killed each year in LX accidents but no reliable data for the number of serious injuries. There is also no comparable data between countries for the type of victim – road vehicle driver or passenger, pedestrian, train passenger etc. There is data for the number of LX accidents per year. Data is also needed for the factors that might be used to scale the risk data. Eurostat publishes some of this data and the United Nations Economic Commission for Europe also issues useful statistics. In practice it was not easy to find comprehensive and reliable data. Different sources were inconsistent; the most authoritative was used but even then private information available to the author suggests even that is not accurate. Data for European countries that are not members of the EU is even more difficult to find. Where it was possible to obtain an almost complete dataset for a country but one number was unavailable, the missing value was estimated by assuming that the country conformed to the European average. For example,



PROCEEDINGS 49/208

for one country it was not possible to find the number of passenger car km per year. It was assumed that this would be: (number of cars in that country) x (total passenger km in all other countries)

(total number of cars in all other countries)

Data for 26 European countries was included in the research.

Factors that might contribute to risk

The table below lists some of the factors that might influence the level of risk at level crossings. Some of them cannot be measured directly so the second column lists readily-available data that might be used as a surrogate for the risk factor. Risk factors Readily available data number of LX Ideally the SI would be broken down according to

different types of LX. ERA’s working group is developing such a classification but this is still not complete. This note considers only the total number of LX of all types. Symbol: N

intensity of use by trains – ideally the frequency with which trains pass each LX

A useful measure of the frequency of use by trains is the number of train km per year divided by the length of track in km. This is a measure of the average number of times that a train passes any point on the network each year. Symbol: T

intensity of use by road vehicles – ideally the frequency with which road vehicles pass each LX

Data is readily available for each country for the numbers of cars, buses and commercial vehicles in use, for the numbers of passenger km by car and by bus and the number of vehicle km for freight vehicles. For purposes of this paper, two measures of the intensity of use of the roads are tested: • the number of cars on the roads divided by the

total length of all roads in the country: Symbol R1 • the number of car passenger km divided by the

total length of all roads in the country: Symbol R2 These measures include many assumptions, such as that the number of bus and freight movements is proportional to the number of car movement, and that the number of car km is proportional to the number of car passenger km (ie the average number of people in a car is the same in every country).

level of risk accepted by road users – ideally a measure of the propensity of national road users to breach road safety laws and therefore to abuse level crossing safety.

Much of the risk of LX arises from abuse by road users who fail to obey warnings and signs. It is believed that the willingness to do this varies with national culture. For the purposes of this paper, two measures of the propensity of national road users to breach road safety laws are tested: • number of people killed on the roads per year

divided by the population of the country: P1 • number of people killed on the roads per year

divided by the number of car passenger km: P2

Table 1: Definitions of data used for scaling factors.



PROCEEDINGS 50/208

There are many different SI that might be constructed even from that small number of factors. This paper uses 14 different Safety Indicators for illustration; more could be defined:

Description Definition

SI 1 Number of fatalities per year per LX Fm / (Nm)

SI 2 Correction for train intensity Fm / ( Nm x Tm)

SI 3 Correction for train and road intensity Fm / ( Nm x Tm x R1m)

SI 4 As SI 3 with different measure of road intensity Fm / ( Nm x Tm x R2m)

SI 5 Correction for train and road intensity and road driver behaviour Fm / ( Nm x Tm x R1m x P1m)

SI 6 As SI 5 with different measure of road driver behaviour Fm / ( Nm x Tm x R1m x P2m)

SI 7 As SI 5 with different measure of road intensity Fm / ( Nm x Tm x R2m x P1m)

SI 8 As SI 6 with different measure of road intensity Fm / ( Nm x Tm x R2m x P2m)

SI 9 Correction for road driver behaviour Fm / ( Nm x P1m)

SI 10 As SI 9 with different measure of road driver behaviour Fm / ( Nm x P2m)

SI 11 Correction for train intensity and road driver behaviour Fm / ( Nm x Tm x P1m)

SI 12 As SI 11 with different measure of road driver behaviour Fm / ( Nm x Tm x P2m)

SI 13 Correction for road intensity Fm / ( Nm x R1m)

SI 14 Simple ratio of LX fatalities per year to road fatalities per year, no allowance for number of LX or intensities

(No of LX fatalities) (No of road fatalities)

Table 2: Definition of Safety Indicators tested

A similar set can be constructed using the number of accidents, Am, instead of Fm. In practice it is convenient to adjust each SI so that the average across Europe is 1. The values for each country then represent the amount by which the risk in that country varies from the European average. This allows the different SI definitions to be compared by inspection.

Results

The calculated Safety Indicators are shown in Annex 1. 1. The results using the number of accidents (Am) were similar to those for number of fatalities (Fm) but

showed consistently greater spread. This probably reflects greater variation in definitions between countries. The accident-based results have been omitted; the conclusions would be the same if they were included but are less clear because of that greater scatter.

2. The countries cannot be identified from the data; they are in a random order. This is necessary because

some of the data was provided in confidence to the author.



PROCEEDINGS 51/208

3. Three different measures of spread are shown: the standard deviation of the SI; the ratio of maximum to

minimum value of the SI and a root mean square measure that is weighted according to the size of the country. A single overall spread is calculated by multiplying these three together.

4. The best predictor of safety risk is SI 14 – the number of road fatalities in the country per year. The next

best is SI 10, which is (number of level crossings) x number of people killed on roads per year per passenger km. The third best is SI 1, which is the number of level crossings.

5. The apparent relative safety performance of countries depends critically on which SI definition is adopted.

For example: - 17 of the 26 countries are above the European average according to at least one definition of SI and

below the European average according to another - 7 different European countries have the best safety performance, depending on which definition of SI is

adopted - 6 different European countries have the worst safety performance, depending on which definition is of

SI adopted

Conclusions

1. On the availability of data:

1.1 Reliable and consistent data is hard to find, especially for countries outside the EU. This may improve when the first set of Common Safety Indicator (CSI) data is published by ERA, although the problems of inconsistent definitions might remain.

1.2 There is a wholly inappropriate concern about secrecy. Safety improvements come when safety performance is exposed to public scrutiny; the railway sector in many countries hides its safety performance. This too may improve with the publication of CSIs.

1.3 Even with those concerns, there is a wealth of data available and properly designed CSIs could be used.

2. On the predictors of LX accident rates:

2.1 The results indicate clearly that the dominant factor that determines LX accident rate is the behaviour of road users.

2.2 The number of LX in the country is also relevant but there is little evidence that the intensity of use of the railway is significant.

2.3 Given these conclusions and the observation that most of the victims of LX accidents are road users, there is powerful support for the argument that LX safety is primarily a roads issue and not a rail issue. Action to improve safety may require cooperation between rail and road authorities but these results provide support for an argument that the initiative and funding should come from the roads authorities.

3. On the calculation of Common Safety Indicators:

3.1 There are many different ways of calculating a Safety Indicator that allow comparison between countries. This paper has illustrated this with 14 different definitions, all of which are plausible but which reach very different relative performance indications.

3.2 If indicators of this type are to be used as the basis of legally binding safety targets, there must be a robust and defensible reason for selecting the data used to measure safety (or lack of safety) and the data used to scale that data to allow for the scale of rail operations in that country.



PROCEEDINGS 52/208

Acknowledgements and Disclaimer

Some of the original preparatory work for this paper was carried out under contracts funded jointly by UIC Safety Platform and the UK Rail Safety and Standards Board.

The views expressed in this paper are solely those of the author and do not necessarily represent the views of any other organisation or person with whom the author has a professional involvement.

Annex: Table of Safety Indicators and calculated sp reads

Country SI 1 2 3 4 5 6 7 8 9 10 11 12 13 141 0.4 0.8 1.8 2.5 2.2 1.9 3.0 2.7 0.5 0.4 1.0 0.9 0.9 0.22 0.6 0.6 0.7 0.6 0.9 1.6 0.8 1.4 0.7 1.3 0.7 1.3 0.7 0.73 2.4 6.5 4.2 4.4 4.2 8.1 4.5 8.6 2.4 4.6 6.6 12.7 1.5 2.04 1.9 2.3 3.5 4.3 2.7 3.2 3.3 4.0 1.5 1.8 1.8 2.1 3.0 1.05 0.8 1.1 0.6 0.7 0.6 0.4 0.6 0.4 0.8 0.5 1.1 0.7 0.4 1.16 4.8 16.6 46.5 46.2 80.2 19.9 79.7 19.8 8.3 2.1 28.6 7.1 13.5 1.57 5.0 9.7 39.4 40.2 32.8 36.6 33.5 37.4 4.2 4.7 8.1 9.0 20.4 2.58 2.2 1.5 1.8 1.6 1.7 2.7 1.5 2.4 2.1 3.4 1.4 2.3 2.6 1.39 1.8 2.1 2.6 3.1 2.0 2.0 2.4 2.4 1.4 1.4 1.6 1.6 2.2 3.610 0.6 0.4 0.2 0.2 0.3 0.5 0.3 0.5 0.8 1.3 0.5 0.9 0.3 0.711 2.7 3.6 1.1 1.3 1.1 1.3 1.3 1.5 2.7 3.3 3.5 4.3 0.8 0.612 0.5 0.3 0.2 0.1 0.3 0.5 0.3 0.5 1.0 1.7 0.6 1.0 0.3 0.413 2.4 5.9 54.1 7.8 38.0 34.8 5.5 5.0 1.7 1.6 4.2 3.8 22.1 0.714 1.3 2.2 2.6 3.7 1.8 1.3 2.6 1.8 0.9 0.7 1.6 1.1 1.5 1.115 0.4 0.9 2.9 4.0 1.3 1.0 1.9 1.4 0.2 0.1 0.4 0.3 1.3 0.216 0.2 0.4 0.7 0.6 1.4 2.3 1.2 1.9 0.4 0.7 0.8 1.4 0.4 1.117 0.6 0.4 0.3 0.2 0.4 0.7 0.4 0.7 0.9 1.7 0.6 1.0 0.4 2.318 0.3 0.1 0.2 0.1 0.3 0.5 0.2 0.4 0.4 0.7 0.2 0.4 0.4 0.519 5.0 11.1 47.9 28.3 22.9 26.2 13.5 15.5 2.4 2.7 5.3 6.0 21.6 1.320 1.3 0.5 0.4 0.5 0.3 0.3 0.4 0.4 1.3 1.2 0.5 0.5 0.9 1.121 0.9 0.4 0.2 0.2 0.5 0.7 0.5 0.7 1.9 2.5 0.8 1.0 0.6 2.122 0.5 1.1 1.8 1.5 2.6 4.6 2.2 3.8 0.7 1.3 1.6 2.8 0.8 2.023 1.1 1.7 3.5 4.6 2.9 2.0 3.8 2.6 0.9 0.6 1.4 0.9 2.4 1.724 0.2 0.3 0.3 0.3 0.6 1.0 0.6 0.9 0.4 0.6 0.5 0.8 0.2 1.425 1.0 0.8 0.6 0.6 0.6 1.1 0.6 1.1 1.1 2.0 0.8 1.5 0.7 0.426 1.2 0.8 0.8 0.8 0.8 1.2 0.8 1.2 1.2 1.7 0.8 1.2 1.1 2.6

Spread: STD 1.45 4.04 16.9 12.3 17.8 10.6 16.4 8.14 1.64 1.21 5.61 3 6.93 0.84Spread: max/min 27.2 112 328 331 276 112 368 104 40.4 31.4 124 39.6 94.2 19.4Spread: weighted 0.37 0.54 0.96 0.92 0.6 0.41 0.58 0.4 0.29 0.26 0.37 0.3 0.58 0.29Overall spread 14.4 244 5320 3741 2967 487 3505 342 19.2 9.99 259 35.2 381 4.8



PROCEEDINGS 53/208


French Safety Improvement policy of Level Crossings

Author(s): Philippe Feltz

Job Title: Programme Manager, Level Crossings

Company: Réseau Ferré de France (RFF)

Country: France

Resume of Speaker

In 1994, Philippe Feltz obtained the degree of Engineer from an ENSI (Ecole Nationale Supérieure d’Ingénieurs). From 1995 until 1997, he was Engineer in charge of the maintenance on a production site of electrical engines. Philippe was Consultant in Maintenance organisation and large projects manager from 1998 until 2003. He joined RFF in 2004, where he started in the audit department, and since 2006 he has been Programme Manager in charge of Level Crossings.

Abstract Presentation of the French enhancement policy of safety at level crossings. This presentation will outline the level crossing population in France, their features, the type of accidents occurring there, as well as the national organisation put in place in 1997, and the subsequent actions foreseen: suppressions, improvement, and prevention operation.



PROCEEDINGS 54/208

Session 3 - Risk Management Chairman: Jürgen Menge Dept for Transport, Rhinelan d, Germany

Level Crossing Accidents in Indian Railways – Risk Reduction Measures Author(s): M.C. Murali Job Title: Chief Safety Officer Company: Indian Railways (Southern Railway) Country: India

Resume of Speaker

M.C.Murali is an Electrical Engineer with a Masters degree in Power Systems and High Voltage Engineering. He also has an Advanced Diploma in Management (Operations Research) After completing his Master’s degree he had a two year stint as Lecturer in Electrical Engineering before joining the Indian Railway Service of Electrical Engineers in 1979. He has put in more than 29 years of service in various capacities in the Divisional units, Rail wheel factory, Centre for advanced maintenance technology, Railway electrification project and Zonal railway headquarters office of Indian Railways. The major areas of activity are operation and maintenance of Electric Locomotives, Electrical Traction system, Train lighting and Air-conditioning, Railway safety, Accident inquiries and Post-accident Management. He has been at the level of Chief Engineer for the past 10 years. Presently, as Chief Safety Officer, he heads the Safety department of Southern Railway - one of the 16 zonal railways of Indian Railways. Prior to this conference, he has presented papers on Railway Safety related topics at the 15th International Railway Safety Conference-2005 at Cape Town, South Africa and 17th International Railway Safety Conference-2007 at Goa, India.

Abstract

On Indian Railways, derailments which once constituted the majority of train accidents have steadily been brought under control over the past quarter of a century, whereas the number of level crossing accidents have remained steady or shown an increasing trend. Now a stage has come when the level crossing accidents are poised to emerge as the major contributor to the number of train accidents. A study of 37 accidents which occurred at unprotected level crossings over a six year period from 2002-2008 on Southern Railway (one of the sixteen Zones of the Indian Railway network) have been analysed in this paper with a view to correlate the theoretical risk factors which are generally identified as contributory to such accidents with the actual situation. The paper details the measures adopted by the Indian Railways to reduce the risk at the level crossings and compares their relative merits. The paper also mentions about the latest measure – limited-use subways - a cost effective and quicker means adopted to expedite the process of elimination of level crossings, with least disturbance to the road and rail traffic.



PROCEEDINGS 55/208

Introduction

By the size of the network, volume of traffic carried and manpower employed, Indian Railways has not many parallels in the world. The entire Railway system is Government-owned and is under a common administration. The network has more than 63,300 route-kms of track carrying more than 750 million tonnes of freight traffic per year and around 18 million passengers per day. Major portion of the network is having broad gauge (1,676 mm) tracks, but some portions are still on metre gauge (1,000 mm) and certain limited stretches on narrow gauge. High density and suburban routes totalling around 28% of the network is electrified. Indian Railways employ around 1.5 million staff in the 16 zonal railways, five production units and other ancillary units including the R&D establishment, training institutes etc.

Quarter of a century back, derailments caused majority of the accidents on the Indian Railways and the number of LC accidents were comparatively smaller. However, over a period of time, on account of improvements made in both track as well as rolling stock, the number of derailments has progressively come down and now we are reaching a stage when the LC accidents are likely to become the majority contributor to train accidents.

The rail-road intersections are either level crossings or road over-bridges/under-bridges, as elsewhere in the world. However, when it comes to protecting the level crossings, the Indian Railways go in for provision of barriers with or without interlocking with signals. Barriers are not automatically operated, but invariably operated by staff provided at the level crossings. There exist a very large number of level crossings on Indian Railways (around 34,000) with all types of vehicles plying across these locations. This would mean that a train on an average encounters one level crossing at every 1.8 kms. It puts the loco drivers under a lot of stress. Majority of these level crossings on Indian Railways (around 18,000) are not equipped with any protective systems such as barriers/gates or train approach warning devices. Combined with this, the road vehicle population is increasing phenomenally. Further, with new generation air-conditioned vehicles with closed windows, music systems etc. the road vehicle drivers do not get adequately warned by the whistling sound of the approaching trains.

The system of risk evaluation at the level crossings is by and large based on a figure of Train-vehicle units (TVUs) obtained at the particular crossing, which is the product of the number of road vehicles and trains passing through the location in a period of 24 hours. This is the basic figure which is used for determining the priority for up-gradation of a particular level crossing. Even though the TVUs provide a simple logical means of risk assessment, accidents statistics indicate that the figure does not in practice correlate with the occurrence of an incident. There could be several more factors which decide the level of risk at a level crossing and in an effort to identify some of these additional factors, the details of 37 level crossing accidents which occurred on Southern Railway, one of the sixteen zonal Railways in India is analysed.

Train Accidents on Indian Railways

The trend of total number of train accidents on Indian Railways from 1990-91 till 2007-08 along with the category-wise break-up is in Table: 1 below:

Year Collision Derailments LC accidents

Fire accidents

Misc. accidents Total

90-91 41 446 36 9 532 91-92 30 444 47 9 530 92-93 50 414 51 9 524 93-94 50 401 66 3 520 94-95 35 388 73 5 501 95-96 29 296 68 5 398 96-97 26 286 65 4 381 97-98 35 289 66 6 396 98-99 24 300 67 6 397 99-00 20 329 93 21 463



PROCEEDINGS 56/208

00-01 20 350 84 17 2 473 01-02 30 280 88 9 8 415 02-03 16 218 96 14 7 351 03-04 9 202 95 14 5 325 04-05 13 138 70 10 3 234 05-06 9 131 75 15 4 234 06-07 8 96 79 4 8 195 07-08 8 100 77 5 4 194

Table: 1 – Trend of Accidents in IR with Category-wise break-up

The figures generally show a declining trend in the total number. The following graph shows the trend of total number of accidents and the contribution of derailments in the total number. It is evident that both the curves follow each other indicating the predominant effect of derailments in determining the overall number of accidents.

0

100

200

300

400

500

600

90-91

91-92

92-93

93-94

94-95

95-96

96-97

97-98

98-99

99-00

00-01

01-02

02-03

03-04

04-05

05-06

06-07

07-08

Year

Derailments Total

Chart: 1 –Trend of Derailments compared with Total Number of Accidents

But, if we further go into the percentage contribution of different categories of accidents in the total number, it may be seen that more than 80% of the accidents were derailments during the early nineties. This figure has been steadily coming down and now we are at a stage when even though derailments are still the majority factor, their contribution is only around 50%. At the same time, level crossing accidents, which contributed only less than 10% of the total number during early nineties, have steadily increased their percentage contribution over the years and have now reached a level of around 40%.



PROCEEDINGS 57/208

0

10

20

30

40

50

60

70

80

90

100

90-91

91-92

92-93

93-94

94-95

95-96

96-97

97-98

98-99

99-00

00-01

01-02

02-03

03-04

04-05

05-06

06-07

07-08

Year

Derailment LC Accidents

Chart: 2 – Percentage Contribution of Derailments and LC Accidents in the Total Number of Accidents

It is amply evident from Chart: 2 above that before long, the level crossing accidents may take over as the majority contributor to the total number of accidents on Indian Railways. Combined with the fact that majority of loss of human lives and injuries arising out of train accidents are also contributed by the level crossing accidents, means for prevention of level crossing accidents assume paramount importance. With the very large numbers involved and resources required for upgrading the level crossings, it is essential that some kind of risk assessment at these locations is made to get maximum returns in terms of enhanced safety for the investments made.

Analysis of Level Crossing Accidents on Southern Ra ilway

In an effort to identify the contribution of some of the factors generally believed to cause level crossing accidents in real life situations, 37 level crossing accidents which occurred on Southern Railway during the period 2002-2008 were analysed. Out of these, only two accidents occurred at level crossings provided with barriers and manned while the remaining 35 were at barrier-less level crossings.

Train-Vehicle Units

As per the current norms being followed on the Indian Railways, Train-vehicle units (TVUs) obtained at the level crossing is considered as the major parameter to decide its risk level. Broadly, a TVU of 6,000 is considered as the threshold which qualifies an unmanned level crossing for manning. However some other considerations like visibility of the road users can make a gate more vulnerable even with lesser TVUs and qualify it for manning as per the current policy.

The Train vehicle units (TVUs) obtained at the unma nned level crossings where the 35 accidents occurred are plotted below in Chart: 3.



PROCEEDINGS 58/208

0

2000

4000

6000

8000

10000

12000

14000

TV

U

Chart: 3 – TVUs of Level Crossings where 35 Accidents occurred during the period 2002-08

It is interesting to note that in only six cases out of thirty five (19.4%) the TVU is more than the threshold value of 6000 which makes the location eligible for provision of barriers and manning. The chart above also indicates that at majority of the level crossings where these accidents took place, the TVU obtained was not even half the value fixed as the threshold for considering up-gradation of the level crossing. It also remains a fact that there were so many level crossings with TVUs much above the threshold value where no accidents occurred during the same period.

The product of the number of road vehicles and trains crossing at a location gives a logical and mathematically rational figure of risk assessment at a level crossing. This number, at least theoretically, gives all the possible combinations of an accident between a road vehicle and train at the location; higher the TVU, higher the probability of an accident. There has been a theory that this need not necessarily be the case since at a location where a higher TVU is obtained due to higher number of road vehicles, there is a possibility that a previous road vehicle has observed an approaching train and is waiting, which provides an additional warning for vehicles which are following.

In any case, going by the available data on accidents, the figure of TVU cannot be considered as a predominant factor for assessing the risk at a level crossing.

Time of Accident

Another factor which is considered to increase the risk at level crossings is the time of the day. It is generally felt that the vulnerability is high during night hours as compared to daytime. Probably visibility of an approaching train is considered to be better during daytime. The time at which the 35 accidents took place has been plotted in Chart: 4 below.



PROCEEDINGS 59/208

0.00

6.00

12.00

18.00

24.00

Tim

e

Chart: 4 – Time of the day at which the 35 Accidents occurred during the period 2002-08

It may be seen that only in nine out of thirty five cases (26%), the accident occurred between 18.00 hrs and 06.00 hrs i.e. beyond the daylight hours. Even out of these nine, three cases were only marginally beyond the daylight hours. From this real life data, there is no evidence to indicate that the probability of level crossing accident occurring during night time is high. It may also be a fact that the road traffic is considerably lesser during night time as compared to day. Gauge

Generally the important and high density routes are on Broad gauge and some of the lesser utilised sections and branch lines are under Metre gauge. As a matter of policy, all lines are getting converted to Broad gauge. During the period considered for analysis, on an average 30 % of the tracks were on Metre gauge system. The density of train traffic and speed of trains are considerably lower on the Metre gauge system and accordingly the risk of accidents at the level crossings on the Metre gauge tracks is generally considered to be lower than that on the Broad gauge system. But out of 35 accidents, 11 (31%) were on Metre gauge tracks almost exactly correlating with the percentage of Metre gauge tracks on the Railway. Thus there is no reason to consider the level crossings on Metre gauge tracks to be at a lesser risk than those on the Broad gauge system. Visibility to Road Users

As per the provisions in the Motor Vehicles Act of the country, drivers of road vehicles are expected to ensure that no trains are approaching from either side of the track before attempting to cross the tracks at unprotected level crossings. To enable the road vehicle drivers to ensure this, certain minimum visibility distances are prescribed, based on the speed of the trains in the section. In case the minimum visibility is not available due to track curvature or other fixed obstructions, the train speeds at the location are suitably reduced to provide adequate time for the road users to react. Restricted visibility of the approaching train to road vehicle drivers was checked for all the 35 cases. Some restriction of visibility was available only in six out of the 35 cases (17 %). This is also not sufficient evidence to consider it as a predominant risk factor. Previous History

Any level crossing with a history of accident is generally considered to be at a higher risk. But only in one case of the thirty six, there was a previous history of accident at the same location.



PROCEEDINGS 60/208

It is also expected that if an accident has occurred, the road users near the locality will become more cautious and be careful in negotiating level crossings. But this also has not proved correct in the cases that were analysed. There were two accidents on 13.08.06 and 24.08.06 (11 days gap) in two level crossings only 7 kms apart on the same section of track. Similarly there were two accidents on 31.10.06 and 21.04.07 (within 6 months) at two level crossings separated only by 3 kms along the same track. There were also two accidents occurring at the same level crossing on 29.06.07 and 02.03.08 (9 months gap).

From the foregoing analysis of actual accidents, it would appear that there are no predominant risk factors which can be identified with a reasonable amount of certainty to contribute to a barrier-less level crossing accident. The only thing which can be stated with certainty is that the best way to eliminate a barrier-less level crossing accident is to eliminate the level crossing itself.

That brings us to the discussion of risk reduction measures or measures to eliminate barrier-less level crossings.

Risk Reduction at Level Crossings

Up-gradation of Level Crossings

As far as Indian Railways is concerned, the order of up-gradation of level crossings for risk reduction is as follows:

1. Unmanned to manned – Full length barriers/swing gates are provided for closing the level crossing against movement of road vehicles and these are locally operated by staff posted at the location. These staff are also provided with communication facility with the nearest station so that the Station master can get a confirmation from the gatekeeper before permitting a train into the section. However, at this stage there is no interlocking of the gate position with the signals.

2. Manned to Interlocked - At the next stage of up-gradation, the gate position is interlocked with signals and unless the gate is closed and locked, the signals cannot be taken off for movement of trains.

3. Replacement with Road Overbridge/Underbridge – The level crossing is replaced with a grade separator, thereby totally eliminating the problem

At the first step, a large amount of risk still exists since the process of closing the barrier/gate and communication is to be done by the gatekeeper and there could be human failures like confirming closure without actually closing or opening the gate after confirming but before the train has passed the level crossing. Further, the Station master concerned could also commit lapses by sending a train without getting confirmation of closure from all the level crossings in the section. Thus risk reduction by this step is again dependent on the human interface involved in the operations. During the period 2002-2008, two cases of accidents occurred at such manned level crossings (in addition to 35 cases at unmanned level crossings) on account of lapses by the gatekeepers.

The next step of providing signals interlocked with the gate position is a more fool-proof and safe method, though much more expensive to install operate and maintain as compared to the first step. Though there were no accidents at such level crossings on Southern Railway during the period analysed, there have been cases on other parts of Indian Railways. Hence this arrangement is also not considered totally risk free.

The step of elimination of level crossings by providing a Road underbridge/overbridge is not only a very expensive method, but it is associated with several other problems related with routing the connecting roads, complicated co-ordination works with several other agencies, land acquisition and long gestation periods.

In an effort to reduce the cost of construction of grade separators to replace level crossings and complete the construction activities in a limited period of time, with least amount of disruption to road and rail traffic, a new concept called “limited-use subways” have been tried at various locations on the Indian Railways. This is mainly directed at closure of some of the level crossings which are not used by bigger vehicles and where the track is already located on an embankment with height of around 3m to 5m from adjoining ground level. At



PROCEEDINGS 61/208

some of the locations, it may also be possible to raise the track level to a small extent for accommodating this arrangement.

The construction process involves the following steps:

1. Pre-casting of rectangular cement concrete box sections at a convenient location

2. Transporting these Concrete sections to the site of construction

3. Selecting a suitable date and time to block the road and rail traffic for a period of about 6-8 hours

4. Removing the track and formation for the required width after blocking the road and rail traffic

5. Excavating the level crossing location for suitable depth to accommodate the box sections

6. Placing the box sections in place after preparing the ground

7. Restoring the formation and re-laying track

With proper planning, it has been possible to execute the actual subway construction within the traffic block period of around 6 – 8 hours. This is in sharp contrast against the construction of conventional Road overbridges/underbridges which takes several months of activity, imposing serious restrictions on movement of rail and road traffic at the location. The cost involved is also restricted to around Rs. 5 million (US $ 125,000) which is less than one-tenth of the cost of constructing a conventional Road overbridge.

However, construction of these subways has the following limitations:

1. Adequate bank height should be available at the location so that the existing road level need not be lowered, which may cause water-logging problems

2. Only limited headroom is available for the road vehicles and only those locations where such vehicles ply can be considered. Designs have been made for providing headroom of 3.66 m and 2.75 m. This is against 4.67 m headroom provided normally at track crossings with overhead traction wires.

3. Planning for execution has to be done meticulously and all the handling machinery kept ready so that the track and road can be restored within the limited period of traffic block

Summarising, the various options for rail-road intersections are compared in Table: 2 below from the point of view of Capital cost, Running cost and Effectiveness:

Type of track crossing Initial Cost Operating Cost Effectiveness

Unmanned LC Low Low Poor

Manned Non-signalled LC Medium Medium Medium

Manned Signalled LC High High Good

Overbridge/Underbridge Very High Medium Very Good

Limited-use subway High Medium Very Good

Table: 2 – Relative comparison of Level Crossing and Grade Separator Options



PROCEEDINGS 62/208

Works other than up-gradation

Other than the up-gradation works as above, there are certain steps which could in some way help in reducing the level crossing accidents, as mentioned below. These are not very expensive to implement, but they make the road vehicle users aware of the impending dangers at level crossing.

Sign boards: Warning signboards, preferably retro-reflective or luminous type are provided on the approach to level crossings so that the road vehicle drivers are adequately warned that they are approaching a level crossing. In addition, sign boards are also provided along the track on the approach to level crossings for the train drivers to sound their horns and alert the road users who may be approaching the level crossing.

Speed Breakers: Speed breakers are provided on the road close to the level crossing so that the road users get a further physical prompt to stop the vehicle and look out for approaching trains.

Road Approach and Surface: A properly graded approach road having an even surface can help in the road vehicle to pass through smoothly without getting stalled/stopped. In addition, there has to be an adequately level road surface just before the tracks where the road vehicle driver can stop his vehicle and look out for approaching trains.

Good Visibility: Providing good visibility of the approaching trains to the road users is another area which can reduce the risk at barrier-less level crossings. Periodic checks at these locations and removal of bushes, tree branches and other obstructions can help.

Warning Devices: Trials are being conducted with a “train actuated warning device” which can provide an audio-visual warning to the road users at barrier-less level crossings. This device gets actuated by an approaching train at a distance of around 2 kms and continues to remain actuated till the train passes the level crossing. However, the onus is still on the road user to stop or cross the tracks.

Publicity: Educating the road users about the rules to be followed while coming across level crossings and the dangers of trying to cross tracks on the face of an approaching train by means of short films, newspaper advertisements and other campaigns may have some effect on the behaviour of the road vehicle drivers at level crossings.

Enforcement: Along with educating road users, some amount of random checks on the observance of correct procedure at barrier-less level crossings by the road vehicle drivers and counselling/punishing the errant persons can have some salutary effect on the behaviour of the road users.

Conclusion

Study of 37 real life accidents at level crossings indicates that there is no appreciable correlation between the occurrence of the accidents and some of the identified and perceived risk factors associated with level crossings. However, one confirmed fact is that the propensity of an accident occurring at an unguarded and barrier-less level crossing is far higher than that at those provided with protective barriers. Thus one major risk reduction measure will be to eliminate the unguarded level crossings by upgrading them. But at the same time, if the level crossing should exist in whatever guarded form, there is always a possibility of an accident occurring. So, to totally eliminate a level crossing accident, the level crossing itself should be eliminated. Considering the cost factor involved in eliminating the large number of level crossings by replacing them with grade separators, the relatively new concept of “limited-use subways” can help. Till such time the level crossings are eliminated, several simple risk reduction methods to make the life easier for the road-user can be employed effectively and such accidents avoided.



PROCEEDINGS 63/208


Safety inspections at Finnish level crossings

Author(s): Veli-Pekka Kallberg

Job Title:

Company: VTT

Country: Finland

Abstract

There are 4,430 level crossings in Finland, 740 of which are equipped with automatic half barriers and 100 with flashing lights and bells. Most level crossings (3,600) are passive and have only crossbucks and sometimes advance warning signs for road users. The annual number of level crossing accidents is approximately 50 and result in close to 10 fatalities per year. To improve the safety of level crossing the Finnish Rail Administration (RHK) has since 1999 commissioned VTT to carry out systematic safety inspections at all level crossings on given main railway lines. Since then VTT has inspected approximately 400 level crossings per year so that by October 2007 the total number of inspected level crossings is 3,495. The inspections carried out so far cover approximately 95% of level crossings on national railway lines. This paper describes the methods used and results achieved.



PROCEEDINGS 64/208

Session 4 - Risk Management Chairman: Aidan Nelson, Community Partnerships Ltd. , UK

Proposition of a New Level Crossing Protection Syst em

Author(s): Philippe Feltz

Job Title: Programme Manager, Level Crossings

Company: Réseau Ferré de France (RFF)

Country: France

Resume of Speaker

In 1994, Philippe Feltz obtained the diploma of Engineer from an ENSI (Ecole Nationale Supérieure d’Ingénieurs). From 1995 until 1997, he was Engineer in charge of the maintenance on a production site of electrical engines. Philippe was Consultant in Maintenance organisation and large projects manager from 1998 until 2003. He joined RFF in 2004, where he started in the audit department, and since 2006 he has been Programme Manager in charge of Level Crossings.

Abstract Non-guarded level crossings (NGLC) are set up on less busy roads and paths, often located in rural areas. Surely this category of level crossings (LC) produces comparatively few accidents relative to their number. Even so, for example in France they cause 21% of the total train-vehicle collisions (cf. figure 1). Most of the time, these collisions cause severe human (dead, injury) and material (derailment, traffic perturbation, etc.) damages. Aware of the problems caused by NGLCs (LCs with St. André cross), RFF the French railway infrastructure manager, intends to improve the safety of these LCs by leading the two main actions in parallel in the period 2012 – 2015: o the first action aims to eliminate as many as possible NGLC (at least 50%) by creating alternative ways

towards another LC or better, towards an existent under/over pass o the second action consists of equipping NGLC with automatic protection. So far, when a LC is converted to

automatic operation electricity has to be supplied to the site. In addition the main costs of these arrangements are due to the use of underground cables in order to link the train detection devices to the LC control system, and to carry the electricity. The idea is thus to design an automatic LC protection system easy to set up, without use of cables and without being linked to the electricity network. The design of this new system is the aim of a two-year project called SECUPASS initiated by INRETS-ESTAS, in collaboration with RFF, and performed by a group of engineering students of EC-Lille.



PROCEEDINGS 65/208


A Risk Assessment framework for Road-Rail Level Crossings: Application to Moroccan Level Crossings

Author(s): Abdelaziz Berrado, Abdelghani Cherkaoui, El-Miloudi El-Koursi,

Moha Khaddour

Job Title: Assistant Professor of Engineering Management

Company: Al Akhawayn University in Ifrane (AUI)

Country: Morocco

Resume of Speaker

Dr. Berrado is a faculty member of Engineering Management in the school of Sciences and Engineering at Al Akhawayn University in Ifrane, Morocco. Dr. Berrado holds a Ph.D. (2005) in Industrial Engineering from the Ira A. Fulton School of Engineering at Arizona State University, Tempe, AZ, USA, a M.S. in Industrial and Systems Engineering (2002) from San Jose State University, San Jose, CA, USA and a diplome d’Ingénieur d’Etat in Industrial Engineering (2000) from L’Ecole Mohammadia D’Ingenieurs, Rabat, Morocco. His research interests are in the areas of Data Mining, Quality, Reliability and Safety with the emphasis of developing methods for systems' diagnostics, optimization and control. He published several papers in quality engineering and data mining journals and in international conferences’ proceedings. In addition to academic work, he was a senior engineer at Intel.

Abstract

Road-Rail level crossings (LC) have always been a scene for rare but fatal accidents, occurring due to complex interactions between factors arising from the design and operations of level crossings. A first step towards eliminating the causes of these accidents is through understanding and assessing the risks associated with a given level crossing and acting on them. This paper introduces a risk management framework that serves this purpose. The suggested framework involves several activities, including, hazard identification, risk analysis, evaluation, treatment and control. Having explained the suggested framework, this paper illustrates how it can be systematically applied to mitigate risk at a given Moroccan level crossing and how it can be integrated in the global safety management system.

1. Introduction

Railways are regarded as an economic, efficient, environmentally friendly and very safe mode of transport. Their safety is however questionable at LC where the number of fatal accidents has been significant over the years. A major concern is to understand and remove the risks at LC. The subject of risk has increasingly become a point of shared interest between many entities representing different sectors. According to a definition of the



PROCEEDINGS 66/208

United Nations, risk “refers to the expected losses from a particular hazard to a specified element at risk in a particular future time period. Losses may be estimated in terms of human lives, or buildings destroyed or in financial terms”. In this paper we introduce a framework that can be used to build a generic LC risk model which will lead to increase the understanding of risk profiles at LC and will allow for risk based decision making to take place via a structured representation of the causes and consequences of potential accidents arising from the operations of LC. The suggested framework could be used to model risk in other sectors as well and should be more effective when integrated in a global safety management system. The rest of this paper is organized in four sections. In the following section, we introduce the suggested risk management framework and explain its different components. In Section 3, we use a Moroccan LC to illustrate how to tackle risk at LC using the suggested framework. In Section 4, we give an overview about safety management systems and explain where the suggested risk management framework at LC can be integrated. A conclusion follows.

2. Risk and the Risk Management Process

Risk has been defined both qualitatively and quantitatively with different but converging definitions [3, 4, 15, 16, 22]. Modares [15] defines risk qualitatively as the potential of loss or injury resulting from exposure to hazards. A hazard being considered as a source of danger that is not associated to the likelihood with which that danger will actually lead to negative consequences. Quantitative definitions of risk associate hazards with their probability of nuisance to the people and the environment. For instance in [11], risk is defined to be a set of scenarios (Si), each of which having a probability (or frequency Pi) and a consequence Ci. This quantitative definition to risk aims to estimate the degree or probability of loss related directly to the occurrence of hazards or potential failures of a system. The need for practical assistance in applying risk management in public and private sector organizations, has led to the development of standards on risk management such as The Risk Management Standard [9] and the Australian and New Zealand standard on risk management [1].

The risk management process as set out in the standards consists mainly of five sequential stages, as illustrated in Figure 1, beginning with the establishment of the context within which risk has to be evaluated in order to set both the objectives and scope of the system; this entails an exhaustive and detailed description of the system that is at risk. Having delimited the system, one should identify the potential hazards or sources of risk; in this stage the list of initiating events, Ei, or scenarios of events leading to the undesired outcome is enumerated. Those events include essentially internal and/or external failures of both the technology used and the human force responsible for it. The next stage, usually referred to as risk analysis is reserved for estimating the likelihood, Pi, of the scenarios or events Ei, actually occurring and each scenario’s consequence, Ci, is also estimated. The results of the risk analysis stage are thereafter used to compare and rank the various risk drivers and compute the total expected risk value, R, defined as R=∑Ri where Ri =Pi.Ci is the expected risk value associated with event Ei. In the evaluation stage minor risks may be screened out and more attention will be routed towards risks with highest expected risk value. Risk treatment is the final stage, where action plans are determined in response to the identified risks and mechanisms to control those risks are put in place. It should be noted that this risk management process may well require regular monitoring and review especially when applied with dynamic systems which may evolve over time. Successful risk management requires that all parties who need to be involved at any stage are given adequate opportunity to do so and play an active role in the process and are kept informed of any developments and actions resulting from the process.



PROCEEDINGS 67/208

2.1. Hazard Identification

Hazard identification is often seen as the heart of risk management. The successful accomplishment of this task is critical since if one omits some potential hazards, it could result in severe human loss and infrastructure damage and in a misevaluation of risk. Many hazard identification techniques [20] have been developed in various engineering disciplines. The precursors of these methods were from the Chemical, Aeronautical and Nuclear power industries. Some methods are area specific such as HACCP for the food industry and others that can be applied to almost any system. Preliminary Hazard Analysis (PHA) is defined in [18] as a semi-quantitative analysis that is performed to identify all potential hazards and accidental events that may lead to an accident then rank them according to their severity and thereafter identify required hazard controls and follow-up actions. Several variants of PHA are used, and sometimes under different names for instance Rapid Risk Ranking (RRR) and hazard identification (HAZID). PHA provides an initial overview of the hazards present in the overall flow of the operations of any system, it provides a hazard assessment that is broad, but usually not detailed. PHA often serves as the total hazard identification process when risk is low. In higher risk operations, it serves to focus and prioritize follow-on hazard analyses by displaying the full range of risk issues. PHA can be applied to all subsystems, components and systems. Most of the time, it is performed first, prior to or as an initial step of design, operation, maintenance, and refurbishment. PHA is carried out in four main step beginning with PHA prerequisites where the PHA team is established, the system to be analyzed, its components, boundaries and interactions are defined and described as well as the actors or materials that appear to be the most exposed to risk. Next, all hazards be identified. In the third step of PHA, the consequences of the hazards in terms of infrastructure damage, human injury or loss is evaluated and their frequency is also estimated. Severity and frequency classification may be used instead when historical risk data is not available to make accurate estimations. Finally, the different hazards are ranked in categories based on their severities and frequencies; this may be done through the application of the ALARP principle [3] (Section 1.3). Hazard categorization helps identify which measures and follow up actions should be carried out to remove hazards associated with high risks. Failure Modes, Effects, and Criticality Analysis (FMECA) is a methodology to identify and analyze all potential failure modes of the various parts of a system, the effects these failures may have on the system and how to avoid the failures, and/or mitigate the effects of the failures on the system [18]. FMEA is a predecessor to FMECA. The C in FMECA indicates that the criticality (or severity) of the various failure effects are considered and ranked. Today, FMEA is often used as a synonym for FMECA. Although FMECA was one of the first systematic techniques for failure analysis, it is not able to identity complex failure modes involving multiple failures within a subsystem, furthermore it has a limited examination of human error and external influences. FMECA remains the most widely used reliability analysis technique in the initial stages of product/system development. A Hazard and Operability (HAZOP) study [12] is a structured and systematic examination of a planned or existing process or operation in order to identify and evaluate problems that may represent risks to personnel or equipment, or that may prevent efficient operations. HAZOP was initially developed to analyze chemical process systems, but has later been extended to other types of systems and also to complex operations and to software systems. HAZOP is a qualitative technique which uses special adjectives (such as "more,""less," "no," etc.: being a unique feature) combined with process conditions (such as speed, flow, pressure, etc.) to systematically evaluate deviations from normal conditions. HAZOP also ranks risk based on severity and likelihood and is best suited for the identification of safety hazards and operability problems of continuous process systems, especially fluid and thermal systems and also to review procedures and sequential operations. A major limitation of HAZOP and of the techniques that we introduced thus far is that they focus on one-event causes of deviations. Multiple-phase failures or hazards due to complex interactions of simple event have to be identified based on the hazards previously identified. Several tools are available for this purpose: A Bayesian Networks is a directed acyclic graphical representation of the joint probability distribution for a set of discrete variables. To each variable A is attached the conditional probability of A given the parents of A. The graphical representation makes Bayesian networks a flexible tool for constructing models of causal impact between events, in particular when the causal impact has a random nature. Bayesian Networks can be used to model hazards that are the result of complex interactions of simple event. Cause & Effect analysis provides a structured way to think through all possible causes of a problem, this tool consists of constructing fishbone diagrams [10], and has been successfully used to mitigate quality problems.



PROCEEDINGS 68/208

A Reliability Block Diagrams (RBD) [Lemis] are used to model how the components (represented by "blocks") are arranged and related reliability-wise in a larger system and how they collectively cause system failure. RBDs are used to determine system’s critical components and can also be used to identify multiphase hazards. Event Tree Analysis (ETA) [3] and Fault Tree Analysis (FTA) [13] are hazard identification methods which are able to implement multiple-phase failures, i.e. deal with complex interactions. According to [3], these two techniques give rise to a pictorial representation of a Statement in Boolean logic. ETA uses “forward logic”, beginning by an abnormal (initiating) incident or event and propagate it through the system under study by considering all possible ways in which it can affect the behavior of the (sub) system. After identifying potential accidental events using a PHA, a HAZOP, or some other technique, ETA helps identify all potential accident scenarios and sequences in a complex system. ETA generates qualitative descriptions of potential problems as combinations of events producing various types of problems from initiating events. It also produces quantitative estimates of event frequencies and relative importance of various failure sequences and contributing events. This enables giving recommendations for reducing risks and evaluating their effectiveness. ETA is however limited to one initiating event and can easily overlook subtle system dependencies. On the other hand, FTA uses backward logic, starting from a top event (a potential accident of interest) to seek all the ways it can happen. The analysis proceeds by determining how the top event can be caused by individual or combined lower level failures or events. The causes of the top event are connected through logic gates. FTA generates qualitative descriptions of potential problems and combinations of events causing specific problems and also quantitative estimates of failure frequencies, and relative importance of various failure sequences and contributing events. FTA is the most commonly used technique for causal analysis in risk and reliability studies, it has, however, a narrow focus since FTA zooms on one specific accident; furthermore significant expertise is required for quantification of frequencies.

2.2. Risk Analysis

Risk Analysis consists of the estimation of the frequency and consequences of each accidental event. The frequency may be estimated based on historical data of previous incidents, FTAs or expert judgment. Consequence analysis identifies both immediate consequences and those that are not apparent until sometime after the accidental event. All potential event chains following an accidental event must be identified and described. Consequence analysis may be conducted using event tree analysis, simulations or can be derived from historical data. Cause-consequence analysis [6] is another technique for consequence analysis which explores system responses to an initiating "challenge" and enables assessment of the probabilities of unfavorable outcomes at each of a number of mutually exclusive loss levels. This technique provides data similar to that available with an ETA; however, it offers two advantages over an ETA- time sequencing of events is better portrayed, and discrete, staged levels of outcome are analyzed. It is important to include all consequence categories, include for the case of LCs, rail company personnel, passengers, the environment (road side of LC), the economic impact, operational consequences and rail company reputation. Losses may be estimated in terms of human lives, or buildings destroyed or in financial terms” [5, 21, 19, 34]. In the absence of data, one can adopt an ordinal scale for hazard frequency classification and consequence or severity classification. Tables 1 & 2 give possible classifications for hazard frequency and consequence.

2.3. Risk Evaluation

If all the consequences and frequencies of hazards have been identified then quantitative definition of risk can be used to estimate risk:

R=∑Ri where Ri =Pi.Ci (1) In the risk evaluation step, the existing risks are classified and decisions are made regarding the tolerability of the existing risk. Risk tolerability is generally a complicated and multifaceted issue which raises philosophical questions from several angles. Several principles can be used to determine the acceptable risk: The precautionary principle [17] is a moral and political principle which states that if an action or policy might cause severe or irreversible harm to the public, in the absence of a scientific consensus that harm would not ensue, the burden of proof falls on those who would advocate taking the action.



PROCEEDINGS 69/208

Score Frequency Class

1 Very unlikely

2 Remote

3 Occasional

4 Probable

5 Frequent

Table 1. Hazard Frequency Classification

Score Severity Class

1 Minor

2 Major

3 Critical

4 Catastrophic

Table 2. Consequence/Severity Classification

GAME or GAMAB meaning “globally at least equivalent” [8], can be applied when looking at either individual or collective risk. This criterion is based on the requirement that the total risk inherent in any new system must not exceed the total risk inherent in comparable existing systems. It is assumed that the risk level of existing systems can be assessed (e.g., using existing statistics). The respective risk levels of an existing system and a new system can only be compared if both systems have comparable performance characteristics and operating conditions. MEM (minimum endogenous mortality) [8] requires that the total risk from all technical systems affecting an individual must not exceed minimum human mortality (2E-4 deaths per person per year). ALARP principle [8] ensures that the risks of any system with serious consequences in terms of human loss and injuries, is kept to a level which is As Low As is Reasonably Practicable. ALARP defines three risk levels:

• Intolerable Risk, which cannot be justified or accepted, except in extraordinary circumstances • Tolerable Risk, which can be accepted only if risk reduction is impractical or if the cost or risk reduction

greatly exceeds the benefit gained • Negligible Risk, which is broadly acceptable and does not require risk mitigating measures

If risk is determined to be at the intolerable level, measures must be taken to reduce it immediately to a tolerable level. If risk is found to be at tolerable level, risk mitigating measures should still be applied, provided that a cost benefit analysis is in favor of it. Table 3 illustrates a risk classification matrix based on ALARP principle.

Frequency/ Consequence

1 Very Unlikely

2 Remote

3 Occasional

4 Probable

5 Frequent

Catastrophic Critical Major Minor

Negligible Risk

Tolerable Risk

Intolerable Risk

Table 3. Hazard Categorization based on ALARP principle

2.4. Risk Treatment and Control

Risk treatment is the process of selecting and implementing measures to reduce see remove the risks. Having identified all sources of risks, one will need to prioritize risk treatment actions and target high risk before low risk while maximizing the benefit to society. Two major classes of methods are considered while prioritizing risk treatment actions including economic evaluation and social evaluation. Social evaluation is usually used as a prerequisite to the Economic evaluation in decision making as there are a number of factors that cannot be assessed economically. The Economic Evaluation estimates the expected benefits and anticipated costs of control associated with varying degrees of reduction in risk, using monetary criteria which are amenable to quantitative economic analysis. Several types of analysis techniques can be used for economic evaluation of risk treatment alternatives at level crossings including, cost benefit analysis, cost effectiveness analysis and risk benefit analysis [7]. A major component of risk treatment is risk control which consists of putting in place control mechanisms to make sure that risk is permanently removed.



PROCEEDINGS 70/208

2.5. Monitoring and Reviewing the Risk Management P rocess

Monitoring and review of the risk management process is a means to make sure that the actions taken effective and that the procedures adopted and information gathered throughout the process were appropriate. It should be noted that systems are dynamic which means that they may get exposed to new risks as they evolve over time, reviewing and monitoring enable keeping track of the changes that systems may undergo.

3. Risk Assessment for Level Crossings: Application to a Moroccan Level Crossing

3.1. Description of the system under study

A level crossing (LC) is an intersection between the road and the railway that allows vehicles of any type to pass through it. The “danger zone” is the area of the intersection in which a collision between the incoming train and LC road users (vehicles and pedestrians crossing the LC) can take place. LCs differ in the protection they offer users, their degree of usage, and in the speed and frequency of the trains that pass over them. LCs are categorized into active crossings where the road user is given a warning of incoming train or passive crossings where no warning is provided, the responsibility being on the road user or pedestrian to determine whether it is safe to cross the LC. Moreover, active LCs can be split into two major subcategories i.e. manual and automatic LCs. In Morocco, the only type of active LC used is the manually controlled full-barriers (MCB) which will serve as the basis of our risk assessment study. The Moroccan LC studied is composed of two rail tracks, and is crossed by a two-way road. The LC is operated by a LC keeper who is responsible for lifting and lowering the mechanical full-barriers and also for alerting the different LC actors of the presence of danger at the LC.

Technical characteristics of the Moroccan Level Cro ssings

The Moroccan national railway organization, ONCF, classifies its LCs according to two criteria, namely LC moments and their location. The LC moment corresponds to the number of trains and vehicles (cars and motorcycles) that pass through the LC in a 24 hours period:

LC moment = [Number of trains / 24h] * [Number of Vehicles / 24h] (2) The second criterion, which is related to the location of the LC, corresponds to the visibility of the incoming train by the vehicles drivers. In fact, ONCF defines a sufficient visibility when a person being at 5 meters from the nearest rail track and whose eye is at one meter from the ground sees the complete locomotive (railway engine used to tow railway cars), moving at the maximum authorized speed, for a period of 20 seconds. ONCF classifies LC with a moment in the interval [2000, 5000] and insufficient visibility as first category. These first category level crossings are manually controlled barriers LC and are the subject of our study. Railway signaling: The railway signals include:

• A metallic announcing panel made out of light-sensitive tapes representing a barrier with the LC number at the top of it. This panel is placed before and after the LC at a distance of 700 m when the authorized train speed does not exceed 120 km/h and at 800 m when this speed is greater than 120 km/h.

• An « S » panel placed at 300 m before and after the LC to remind the train driver that he should whistle to alert both the LC keeper and the vehicles passing through the LC of its incoming.

• White-painted pylons located at least at 500 m before and after the LC Road signalling: There exist two types of road signals: advanced signals and position signals:

• The Advanced Signal is a triangular panel A9 placed at 150 m from the LC which informs the road users that they are approaching a MCB LC and that they should decelerate and be cautious at the LC.

• Position Signals are barriers with tapes of 1 meter length each painted in red and white.

Incoming Train Detection System-Electro-Mechanical Detection: ONCF is using Electro-Mechanical oriented pedal in all Train Detection System (TDS) at manned LC. This automated TDS is composed of pedals placed at the middle of each rail track of the railway 3000 m from the LC. The TDS is directly connected via electrical wires to the LC’s control board and when activated the TDS will trigger both the audible and visual signals at the LC, indicating the direction of the incoming train. These devices are installed in a box located at proximity from the LC Keeper’s shelter and the barriers so that the LC keeper can hear and see it perfectly. When the train passes on the rail track, it will activate mechanically the pedals, then the road signal changes from green to red. The incoming train’s audible announcement can only be turned off if both the LC keeper



PROCEEDINGS 71/208

deactivates the system by pushing on a button on his control board and the pedal is no longer active, train passed the location of the pedals. Entities Involved in the Moroccan manually controll ed full barriers crossings: Several entities may impact the normal operations of the MCB crossings, including the condition of the railway, the condition of the road crossing the railway, the condition of level crossing mechanism, the train detection system, the transmission/communication system, the road signalling, the railway signalling and the level crossing human actors which include the train driver, the level crossing keeper, the road user and the control centre operator. Modelling operational interactions at the LC throug h functional diagrams: Many of the existing hazards at LC may be due to operational failures which can be identified by building functional diagrams representing the LC from different perspective and then identifying operational conditions which may lead to accidents. These functional diagrams give a visual representation of the sequence of events and interactions between the different entities involved in the LC operations and enable a detailed functional understanding of the system. For this purpose we built functional diagrams, for the LC under study, from the perspectives of the different actors in the LC including the LC keeper, the road user, the train driver and control centre operator.

3.2. Hazard Identification at MCB Moroccan LC

In order to identify the complete set of hazards surrounding the MCB LC under study, we considered the different entities involved in the LC and the interactions between them described by functional diagrams. We also reviewed the operational specifications and considered all the environment factors around the LC. We considered the human and LC interface. We identified several hazards that can be classified into one of five categories, namely hazards related to the environment of the LC which effect visibility of LC users, hazards related to technical problems, hazards due to non compliance with standards, hazards due to the human factors, and the fifth category includes all the other hazards. Several sub-categories constitute each hazard category. After several brainstorming sessions, we identified 63 potential hazards along the five hazard categories. The pie-chart in Figure 2 illustrates the distribution of the hazards identified by category. According to this chart, the hazard categories, “Human Factors” and “Technical Problems”, with respectively 37% and 29% of the overall system hazards identified, are the two major hazards that can lead to an accident at the MCB LC. Therefore, a detailed analysis of both categories was needed to understand and identify which actors (people or sub-system parts) are responsible for the majority of them and to state if whether some actions can be undertaken by the appropriate authorities to reduce their impact, as a future step.

Hazards Results by Cause Categories

Visibility9%

Technical Problems29%

Non-Compliance with Standards

15%

Human Factors37%

Other10%

Figure 2. Hazards Identification Results classified by cause categories

3.3. Risk Analysis, Evaluation and Treatment at the MCB Moroccan LC

Since we did not have historical data for risk analysis, we used the frequency and consequence classification described in Tables 1 & 2 to rank each of the 63 identified hazards and then categorized them based on ALARP principle as explained in Table 3. This revealed that 18% of the hazards are considered to have negligible risk, 35% have tolerable risk and they include mainly technical problems related to the train and the TDS. The remaining 47% hazards were associated with the intolerable risk category, and most of them were



PROCEEDINGS 72/208

associated with the human factor and technical problems. The next logical step is to take actions to remove hazards with potential intolerable risk. These actions should target human factors and technical problems.

4. Integrating LC Risk Management Framework into Gl obal Safety Management Systems

Major legislative changes have been undertaken across the European community in the last few years. In addition, there are ongoing technological changes that are occurring. Therefore there is the potential for instability and confusion in the railway industry resulting in an overall increase in accident risk. These changes affect not only the organizational and technical innovations developed with the new systems, but also the new stakeholders and financial arrangements derived from the major changes. Safety management is an important issue in all safety critical sectors including railway industry and regarded as an important means for improving safety culture. A safety management system (SMS) [2] is an organisation’s formal arrangement, through the provision of policies, resources and processes, to ensure the safety of its work activity. An effective SMS helps the organisation to identify and manage risks effectively. It allows an organisation to demonstrate its capability in achieving its safety objectives and in meeting regulatory requirements. A crucial aspect of RU’s and IM’s safety management activity will be the management of interfaces. In many member state railways the new organisational structure will increase the number of interfaces, and hence introduce potentially new types of risk. An organisation faces essentially three different types of risk to its operations, namely internal risks, i.e. those associated with activities and locations for which the organisation is solely responsible, external risks, i.e. those originating from systems, people or organisations and processes that are outside the organisation’s control and shared risks, i.e. risks associated with activities or locations for which there are shared responsibilities rather than sole ownership; to manage such risks the organisations have to ensure that compatible approaches are used. Table 4 shows the proposed eleven elements of the SMS that are divided into two parts: Planning and risk control system and learning system. This organisation of SMS structure should be refined at Stakeholders level and should consider the operation, maintenance and renewal phases of the life cycle [8] of the railway system and lifecycle transition should be explicitly considered in risk assessment activity. The LC risk management framework suggested herein should be integrated in element (5) of a SMS.

Planning and risk control system Learning system

(1) Nature and Scope of Duty Holder’s Business (10) Incident and Accident Reporting and (2) Safety Policy Learning

(3)Organisational structure and Responsibilities (11) Monitoring, Auditing, Corrective Measures (4) Competence, Training and Fitness and Annual reports

(5) Risk Management (6) Safety Assurance (7)Emergency Management

(8) Safety Communication and Information (9) Management of Rules and Standards, including Compliance

Table 4: Structure of Safety Management System

5. Conclusion

In this paper, a framework for risk management at level crossing has been introduced. Furthermore we illustrated how it was applied to a manually controlled full barrier LC in Morocco. We suggested different aspects that should be considered during the system definition phase where we suggested using functional diagrams for modeling operations at LC from the perspective of LC actors. It is a critical part for risk management and specifically for hazard identification where we provided different techniques that can be used; our experience shows that involvement of all stakeholders is a prerequisite to the success to this phase can be comprehensive if all parties, initiating event can be unveiled through brainstorming sessions and FTA can model complex interactions of events that have to the potential to lead to accident. Risk analysis can then be carried out provided that historical LC accident and incident data is available to estimate frequencies and consequences; ETA is the ideal tool for estimating consequences of hazards due to multiple causes. The existing risks are then classified and decisions are made regarding their tolerability, the ALARP principle can serve this purpose. A cost benefit analysis then helps prioritize risk treatment actions that should target



PROCEEDINGS 73/208

intolerable risks. Control mechanisms should be also put in place to assess, monitor and review the risk control actions put in place. We would like to emphasize on the importance of having a database of historical accidents and incidents at LC for the success and efficiency for the suggested framework. Finally we should point out that this risk management framework for LC would be more effective if integrated into the global safety management system of railways.

Acknowledgements

This work is conducted within SELCAT project funded by European commission and we would like to thank all project partners. References [1] AS/NZS 4360. Risk Management. Standards Australia, Sydney; 1999. [2] Bearfield G., Mitra S. and El Koursi E.M. “Guidelines for Safety Management System, D2.2.2/V3.0”,

http://samnet.inrets.fr, (May 2004) [3] Bedford, T. and Cooke, R. “Probabilistic Risk Analysis – Foundations and Methods”, Cambridge Press, (2001). [4] Blanchard, B.S. System Engineering Management. John Wiley & Sons, New York. (1998) [5] Burton, I., Kates, R. W., White, G.F. “The Environment as Hazard”, Second Edition, Guildford Press, New

Yord/London, (1993) [6] Clifton A. Ericson II. “Hazard Analysis Techniques for System Safety”, John Wiley & Sons, (2005). [7] Commonwealth of Australia, “Introduction to Cost-Benefit Analysis and Alternative Evaluation Methodologies”,

(January 2006). [8] EN 50126: “Railway applications—The specification and demonstration of reliability, availability, maintainability

and safety (RAMS).” CENELEC, (1999). [9] “The Risk Management Standard”. Published by the Institute of Risk Management (IRM), The Association of

Insurance and Risk Managers (AIRMIC) and ALARM in 2002. www.theirm.org/publications/PUstandard.html [10] Ishikawa, K., Guide to Quality Control, 2nd rev. ed. Available from UNIPUB/Quality Ressources, One Water

Street, White Plains, NY 10601. Tokyo: Asian Productivity Organization, (1986). [11] Kaplan, S., and Garrick, B.J. “On the quantitative definition of risk, Risk Analysis”, vol 1, no. 1. (1981). [12] Kletz, T. A. “Hazop – past and future”. Reliability Engineering and System Safety, 55:263-266, (1997). [13] Lee WS, Grosh DL, Tillman EA, Lie CH. “Fault tree analysis, methods and applications—a review”. IEEE Trans

Reliab ;R-34(3):194–203, (1985). [15] Modarres, M. “What Every Engineer Should Know about Reliability and Risk Analysis”. Marcel Dekker, NewYork.

(1993). [16] Molak, V. “Fundamentals of Risk Analysis and Risk Management”. CRC Press, Lewis Publishers, Boca Raton, (

1997). [17] Raffensberger C, Tickner J, “Protecting Public Health and the Environment: Implementing the Precautionary

Principle”. Island Press, Washington, DC, (1999). [18] Rausand, M. and A. Høyland. “System Reliability Theory; Models, Statistical Methods and Applications”, Wiley,

New York, (2004). [19] Schnieder, E., Slovak, R. and Wegele, S. “New and Conventional Measures for Quantifying Risk in Rail

Transport”. Journal of System Safety, (February 2005). [20] Stewart MG, Melchers RE. “Probabilistic risk assessment of engineering systems”. London: Chapman & Hall;

(1997). [21] United Nations Disaster Relief Coordinator, “Natural Disasters and Vulnerability Analysis”, in Report of Expert

Group Meeting, UNDRO, Geneva. (9-12 July 1979). [22] Wang,J.X. and M.L. Roush. “What Every Engineer Should Know About Risk Engineering and Management”.

Marcel Dekker, New York, (2000).



PROCEEDINGS 74/208


Corporate responsibility for pedestrian risks at le vel crossings

Author(s): Barry Johnson,

Job Title: Head of Environment

Company: Jarvis Group


Resume of speaker Barry Johnson CEnv BSc (Hon) MSc JP MIEMA MCILT MIOSH (RSP) MBES AMIEEM AMISEA is Head of Environment for the Jarvis Group. Barry acted as Team Leader carrying out environmental due diligence/liability, pre-tender audits and site assessments in Estonia for Eesti Raudtee (Estonian Railways Ltd). Barry was also involved in environmental compliance survey and audits for Baltic Rail Services, working with a team that included the Head of Environment Estonian Railways, Estonian Environmental Consultants, Senior Advisor Sustainable Development World Bank Group and Senior Environmental Specialist International Finance Corporation (IFC). Barry has contributed to developing the environmental element of the Jarvis plc Corporate Social Responsibility (CSR) Report. In 2003, Barry commenced a PhD research with the Open University, working title: Understanding the relationship between corporate risk in the UK Rail Industry and how environmental management systems and corporate social responsibility help reduce this risk.

Abstract

While much has been done to look at accidents involving road traffic and locomotives at crossings, here the focus is on uncontrolled crossings particularly User Worked Crossing and Footpath Level Crossing (UWC and FLC) and the risk to pedestrians. A feature that distinguishes this research from other studies is the focus on users rightfully crossing the tracks rather than on trespassers. The paper explores some issues contributing to the risks including human factors, as well as some potential ways of addressing the problem and communication issues.



PROCEEDINGS 75/208

Session 5 - Safety and Human Issues Chairman: Alan Davies, RSSB, UK

JR East’s Efforts in Level Crossing Safety

Author(s): Takeru Tozawa, Akihisa Harada

Job Title: Signalling Design Engineer

Company: East Japan Railway Company

Country: Japan

Resume of speaker In March 2000, Takeru Tozawa graduated from the Graduate School of Science & Engineering, Keio University where he majored in electronic engineering. In April 2000, Takeru joined East Japan Railway Company (JR East), as a Member of the department in charge of designing and constructing signal facilities. In April 2003, Takeru was transferred to the department in charge of maintenance of signalling facilities. In April 2006, he joined the Electrical and Signal Network System Dept. Since then, Takeru has been in charge of the introduction of new technology, maintenance, and investment for level crossings in general, and is responsible for the 3D laser radar crossing obstacle detector that JR East started to install in 2006.

Abstract This paper describes tendencies and approaches regarding level crossing accidents at East Japan Railway Company (hereinafter “JR East”). The paper also introduces the 3D laser radar (hereinafter “3DLR”) crossing obstacle detector that JR East is actively implementing, a type of detector that is considered effective in preventing crossing accidents. 1. Change in Numbers of Level Crossings and Crossin g Accidents JR East reached its 20th anniversary on April 1, 2007. By grade separating and other improvements, we have reduced the 8,358 level crossings present in 1987 when JR East took over the business of former Japan National Railways to 7,211 as of April 2007. The annual number of level crossing accidents has also dropped from 247 at the establishment of JR East to 32 in FY 2006, approx. one eighth that of FY 1987. That proves the effect of installation of crossing obstacle detectors, which we have actively carried out since establishment.



PROCEEDINGS 76/208

10101010

25252525

21212121

15151515

22222222

20202020

2222

6666

6666

5555

6666

8888

5555

8888

4444

0 20 40 60 80

2006

2005

2004

Num ber of Accidents

Late crossing Trapped by barrierFalling from path B reaking through barrierSlipping O thers

36363636

57575757

26262626

3333

4444

3333

7777

3333

3333

0 20 40 60 80

2004

2005

2006

Num ber of Accidents

C ar M otorcycle Pedestrian O thers

0

1000

2000

3000

4000

5000

6000

7000

8000

90001986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

Numbe

r of Leve

l Crossings

0

50

100

150

200

250

300

Numbe

r of Accidents

C lass 1 level crossing C lass 3 level crossing

C lass 4 level crossing Accidents

Fig. 1: Change of Numbers of Level Crossings and Crossing Accidents Note) Class 1 level crossing: Level crossing with crossing gate and signal Class 2 level crossing: Level crossing where a watchman controls the gate at specific times Class 3 level crossing: Level crossing with crossing signal only Class 4 level crossing: Level crossing with no crossing gate and signal 2. Details of Crossing Accidents Of the crossing accidents in FY 2006, accidents involving automobiles account for 80% of the total(Fig. 2). Most of the collisions with trains were caused by automobiles becoming trapped by the barrier of Class 1 crossings—stalled, running off the path or held in traffic on the crossing when the crossing activates—and by vehicles attempting to cross too late (Fig. 3). Accident rates for every 100 Class 3 and Class 4 crossings reach approx. triple those of Class 1 crossings (see Fig. 4), though the actual number of accidents is small. The rate of such accidents has remained the same over the years. We therefore need to focus on the prevention of automobiles being trapped by the barrier and crossing too late and on improvement and elimination of Class 3 and Class 4 crossings. Fig. 2: Change in Numbers of Accident by Cause Fig. 3: Numbers of Accidents by Cause



PROCEEDINGS 77/208

0

0.5

1

1.5

2

2.5

3

2002 2003 2004 2005 2006

C lass 1 level crossing C lass 3 level crossingC lass 4 level crossing

Fig. 4: Accident Rates per 100 Crossings by Class

3. JR East’s Efforts in Accident Prevention Based on the analysis above, JR East has implemented a variety of countermeasures for accident prevention. Here I will introduce some of them.

a) Elimination of Level Crossings by Grade Separation and Integration of Crossings

The most effective and definite way to prevent crossing accidents is to eliminate level crossings by grade separation and integration of crossings. JR East has been eliminating approx. 20 level crossings per year with the cooperation of authorities in charge of road administration. b) Improvement to Class 1 Level Crossing JR East still has around 810 Class 3 and 4 level crossings in its operation area. We have been improving approx. 15 of them per year to Class 1. Since there is large difference between the accident rates of Class 3 and 4 level crossings and of Class 1 level crossings, we will proceed with that improvement, taking into account the environment and the accident history of each level crossing. c) Installation of Red and White Large Crossing Barrier To prevent vehicles from forcing through the gate after closing and to improve the deterrence ability of level crossings, we are introducing red and white large crossing barriers as a new approach. We install those new barriers upon providing sufficient information to users and police as these barriers are more robust than traditional ones and the red and white color pattern is different from the usual yellow and black pattern used in Japan. d) Installation of Overhead Crossing Signals To prevent too late crossing, we will further introduce overhead crossing signals so drivers are better able to see the signals. We are already using this type of signal in many locations, and we will proceed with further installation of those, giving priority to level crossings with heavy railway traffic (Photo 2). e) Further Installation of Crossing Obstacle Detectors Of the level crossing accidents in FY 2006, approx. 80% were caused by automobiles, with approx. 75% of those

Photo 1: Red and White Large Crossing Barriers

Photo 2: Overhead Crossing Signal

Photo 3: Conventional Crossing Obstacle Detector

Photo 4: 3DLR Crossing Obstacle Detector



PROCEEDINGS 78/208

automobile accidents occurring at Class 1 level crossings. And approx. 85% of such accidents involving automobiles occurred at Class 1 level crossings with no obstacle detectors. Automobiles being trapped by the barrier account for approx. 80% of the automobile accidents at level crossings with no obstacle detector and for 45% of total accidents (Fig. 5). The data shows that crossing obstacle detectors have quite remarkable effect on preventing crossing accidents. JR East has gradually installed obstacle detectors, and almost 2,600 level crossings now have those detectors. In the Tokyo area, installation has been completed at almost 100% of the level crossings where detectors can be installed; so, we are proceeding with further installation in rural areas in order of priority. We are also proceeding with the introduction of a new type of crossing obstacle detector (3DLR type) that is better able to detect obstacles and can be installed in more locations with the same or lower cost than conventional laser or loop coil detectors.

f) Education and Promotion through Crossing Accident Prevention Campaign In addition to improving level crossing safety through physical facility improvement, we are also carrying out promotion of safety through various media. We carry out a crossing accident prevention campaign every year in conjunction with Japan’s traffic safety campaign months. • Education for general drivers at driving schools and license centers on actions to take in case of problems

on the crossing, such as training on evacuation when trapped by the barrier • Requesting cooperation from license centers in educating on and handing out fliers for level crossing

safety to drivers at their license renewal • Making education videos • Increased awareness activities on TV and radio

4. 3DLR Obstacle Detector

Here I will explain the above-mentioned 3DLR obstacle detector that has showed good results among crossing obstacle detectors installed in recent years. a) Principle of Detection by a 3DLR Obstacle Detector

A 3DLR emits pulse laser beams to an object, and measures the distance to the object based on the reflection time of the beams. Adding that data to the position of the laser beam head and the direction the beams are emitted, it identifies the three-dimensional position of the object. It can measure the distance to the object at intervals of up to approx. 10 cm based on the time delay of the reflection of the emitted laser beams. Furthermore, the detector can emit laser beams both horizontally and vertically, and scans the whole crossing path every 0.5 sec. The controller of the 3DLR obstacle detector interprets the group of measured points as a mass, and recognizes the mass as an obstacle. This way, the detector detects the obstacle when something remains on the crossing path for more than a

Fig. 5-1: Numbers of Accidents by Obstacle Involved

Fig. 5-2: Numbers of Accidents by Class of Level Crossing

Fig. 5-3: Numbers of Accidents with and without Crossing

Obstacle Detector

WithWithWithWith WithoutWithoutWithoutWithout

16161616

3333

W ith W ithout

3333

3333

19191919

C lass 1 level crossingC lass 3 level crossing

C lass 4 level crossing

3333

11113333

25252525

C ar BikeBicycle Pedestrian



PROCEEDINGS 79/208

specified time while the crossing warning signal is activated.

b) System Configuration of 3DLR Obstacle

Detector Fig. 6 shows the system configuration of a 3DLR obstacle detector. The detector system consists of a laser radar head installed at approx. 10 m from the level crossing at approx. 5 m high and a controller and a data recorder in a cabinet. The laser radar head and the controller are connected with optical fiber cables for data transmission and metal cables for power supply etc.

c) Advantages of 3DLR Obstacle Detector

i) Better detection ability Conventional opto-isolator obstacle detectors (Fig. 7) sometimes cannot detect a vehicle depending on its size, shape, position or direction it is stopped in. A 3DLR obstacle detector can detect smaller objects because it makes measurement at smaller intervals.

ii) Installable even where conventional obstacle detectors cannot be installed Since a 3DLR obstacle detector needs no light beam emitter/receiver between tracks as an opto-isolator obstacle detector does, it can be installed even where there is no space near tracks.

iii) Easy maintenance, safe work An opto-isolator obstacle detector has light beam emitters and receivers at the height of the rolling stock bogies, and the emitters/receivers often get dirty from passing trains after rain. Since the laser radar head of a 3DLR obstacle detector is installed to a high place, it seldom gets significantly dirty. Maintenance staff also can work safely when checking and cleaning, as they don’t have to enter the track area.

iv) Easy installation The installation work of a 3DLR obstacle detector only involves installation of a laser radar head to a special concrete pole outside of the track area, installation of a controller unit, and connection of cabling between the laser radar head and the controller unit. It does not need installation of many light beam emitters and receivers between tracks and cabling for them as conventional obstacle detectors do. That

1st scanning

2nd scanning

50th scanning

All scanning data

Extracting laser beams reflected from the object X

Z

Y

Fig. 5: Detection by 3DLR Obstacle Detector

Fig. 6: System Configuration of a 3DLR Obstacle Detector

Controller

Laser radar head

Relay (media

converter)

Metal cable for power supply (DC 16V)

Optical fiber cable for data transmission

Metal cable for heater (AC 110V) Media converter

Data recorder



PROCEEDINGS 80/208

means that a 3DLR obstacle detector is remarkably easier to install. Design and work for improving the track or widening of the crossing path can also be considerably simplified, because a 3DLR obstacle detector for the most part needs no relocation of system components.

4. Conclusion Level crossing accidents account for approx. half of the railway accidents of JR East, and those accidents have serious impact on human life and society when they occur. Aiming at zero level crossing accidents, we will strengthen both physical and application approaches.

Fig. 7: Conventional Opto-Isolator Obstacle Detector

Fig. 8: 3DLR Obstacle Detector

Laser beam

Monitoring area



PROCEEDINGS 81/208

Session 5 - Safety and Human Issues

Why do people trespass? Finnish Experiences

Author(s): Anne Silla

Job Title: PhD Student

Company: VTT (Technical Research Centre of Finland)

Country: Finland

Resume of Speaker

Ms. Anne Silla (M.Sc.Tech) graduated from the Technical University of Tampere on May 2005. Since her graduation Anne has been working at VTT (Technical Research Centre of Finland) in their railway safety group. During the time at VTT, Anne has participated in many projects, which have been mainly related to safety and security of railway traffic. Anne’s main research topic while working at VTT has been railway trespassing. In September 2007 Anne was selected to TransportNET programme, which is an academic network for transport education and research offering Early-Stage Fellowships. The fellowship is for PhD studies and enables the fulltime research for PhD. Anne is doing her fellowship in the University of Karlsruhe in Germany and continuing her research on railway trespassing.

Abstract

This study was designed to investigate trespassing behaviour at selected sites to find out what people think about trespassing, why they are doing it and what would make them to stop trespassing. First, sites where trespassing is frequent were explored by a questionnaire survey directed to engine drivers. Based on the results more than 100 locations were specified where trespassing is frequent and from these sites three different locations in eastern Finland close to Lappeenranta were chosen for further analysis. After that the trespassers in these three locations were interviewed. The interviews showed that the main reason for trespassing is taking a shortcut. Most people were trespassing while going shopping, jogging, or on their way to school or work. Thirty-five percent of all respondents trespassed daily or almost daily. It is significant that 67% of all respondents answered that they trespassed at least once a week. Half of the respondents assessed that the trespassing is either completely or fairly safe. Overall, 59% of the respondents considered trespassing illegal, 15% considered it legal and 26% did not know. Given that so many trespassers assumed that trespassing is legal, it is worth considering information campaigns to raise awareness of the illegality of trespassing and the dangers related to it as one form of preventing trespassing. In addition to information campaigns, various physical measures (e.g. overpasses and underpasses, fencing, prohibitive signs) should be considered. It is also possible to strengthen information campaigns by combining them with other measures to have more influence on trespassers' behaviour.



PROCEEDINGS 82/208

The most frequently suggested countermeasures along trespassers included building a fence or an underpass or overpass. In addition to the above measures, enforcement or imposition of a fine, installation of a prohibitive sign and information provided by various means were supported. The suggestions for preventive measures indicated by the trespassers were somewhat related to the locations. The measures thought to be effective at one location were not considered so effective at another. According to the answers in this study, the main factor that determined the suggested countermeasures was distance to the closest official crossing site. Introduction Many studies have argued that trespassing is one of the most important railway safety issues [see e.g. 6, 7, 10]. Based on the Finnish statistics [2], fatalities, which can most probably be assigned to trespassers, constitute 68% of all fatalities caused by railway accidents (for the considered years of 2004–2006). Possible suicide cases have also been assessed but they have been excluded from the above fatalities. Finland is not the only country where such a high proportion of people killed in railway accidents are trespassers. In the European Union more than half of all fatal injuries (excluding suicides) were sustained by trespassers in 2006 [1]. According to Finnish law crossing the railway line is only permitted at sites especially marked for that purpose. The penalty for breaking the law is a fine [3]. Another form of trespassing is illegal walking or loitering in the railway area. But although trespassing is illegal, clear and regularly used footpaths over the railway line can be found at many places. Consequently, it is reasonable to assume that trespassing is frequent. Trespasser interviews are part of a bigger study conducted in Finland. In addition to the survey to engine drivers and trespassers interviews, the trespassers at these sites were counted with cameras equipped with motion detectors to investigate the trespassing behaviour. At the same time with the interviews the questionnaires were sent to households living close to railway lines in order to investigate what people living close to the railway line think about trespassing. They were asked their opinions on possible means to prevent trespass, the dangerousness of trespassing and their awareness about the regulations regarding walking in the railway area. In the last phase of the study related to trespassings three different countermeasures were built to selected sites and the aim was to clarify what kind of effect three different countermeasures – building a fence, landscaping and prohibitive signs – have on the frequency of trespassing. Each of the countermeasures was tested at one site. The questions in the interviews in the present study were based on A Community Problem-solving Guide, which was developed during the community trespass prevention programme in Canada [4]. The programme has been established as an initiative of Direction 2006 and it aims at reducing railway trespassing and crossing incidents and related injuries and deaths. As part of the programme C.A.R.E. (Community, Analysis, Response and Evaluation) problem solving model was developed to help communities in identifying and addressing the underlying cause(s) of trespassing. The model included a template for Trespasser Interviews to collect more detailed information about the trespassing problem is collected to determine its’ underlying causes through problem-analysis. This template was used as an example when developing the questionnaire for the present study. This study aimed to identify the sites of frequent trespassing on Finnish railways, to investigate trespassing behaviour at selected sites, and to explore opinions about possible countermeasures to prevent trespassing. Method The study included three phases. First, sites where trespassing occurs were explored by a questionnaire survey directed at engine drivers. In the survey form, drivers were asked about locations where they have frequently observed trespassers and for their suggestions of potential preventive measures. The survey forms were delivered to all workplaces across the whole railway network in Finland. Thus, the engine drivers could focus on problematic sites in the area where they usually drive trains.



PROCEEDINGS 83/208

The survey form included a map of the area close to the workplace and a table where problematic sites were to be listed. In addition, drivers could refer to problematic sites elsewhere in Finland. The survey forms were delivered to the engine drivers’ mailboxes in their workplaces, making them available to all engine drivers. The second phase foresaw a more detailed analysis of three selected sites. Based on the survey results three locations in the city of Lappeenranta were chosen for further investigation. Trespassers at these sites were counted with cameras equipped with motion detectors. In the third phase, approximately 4 months later, some trespassers at the chosen locations were interviewed. The questions were based on the forms used in the Canadian study shortly introduced above [4]. The interview specifically focused on trespassers' movements in the railway area, their possibilities and willingness to change their routes, how dangerous they think trespassing is, and their awareness of regulations regarding walking in the railway area. In addition, they were asked what would stop them from trespassing. The interviews, which were conducted over 2 days at each of three locations, took place between 7:30 a.m. and 17:00 p.m. In total, 46 out of 108 trespassers were interviewed. Some trespassers were not interviewed for of the following reasons: they were identified more than once, some of them were speaking on their mobile phone, interviewers were too occupied, or trespassers were too busy. Results Survey of Engine Drivers In total, 1,500 survey forms were distributed to engine drivers. Ninety-six forms were returned that included 404 problem sites (Table 1). The response rate was only 6%. However, this was not considered a major problem regarding the representativeness of the results, because at least one survey form was returned from each workplace. Therefore it was presumed that at least the worst problem sites on the railway network were identified. Because of a substantial overlap among identified sites, approximately 100 unique locations could be identified. Trespassing seems to concentrate near big cities where the population density is high and rail traffic is dense. In addition, trespassing was seen as a communal problem. Based on the returned survey forms it was possible to identify some locations where people from all the age groups were frequently trespassing. At these locations people do not necessarily consider trespassing as wrong and they may not understand the risk incurred by it.

Table 1: Number of Workplaces of Engine Drivers, Number of Returned Survey Forms and Number of

Reported Problem Sites by Area. The survey results were used to choose candidate locations for the next phase of the study. The most suitable locations for further investigation were identified by preliminary site observations. A total of more than 10 locations from the capital area and the Lappeenranta area were selected. The final selection criteria included the following: (a) it is possible to execute measurements with the help of cameras with motion detectors, (b) the amount of trespassers is relatively high and (c) the legal rail crossing site is located less than 500 m from the trespassing location.



PROCEEDINGS 84/208

The most suitable sites for further investigation were all found in the city of Lappeenranta. The Lappeenranta area is challenging, because the tracks divide the city into two parts. Based on preliminary site observations in Lappeenranta over a 4 km stretch of track, a total of 12 trespassing locations were found of which three were selected. At each location the official route is no more than 300 m away. Interview Trespassers at the three selected locations were interviewed. The reasons for trespassing were asked with an open question rather than alternatives, and the interviewees were able to indicate as many reasons as they wanted. Although the official routes were not more than 300 m away at each location, 80% of interviewees indicated that the most common reason for trespassing was that the route was the shortest and fastest alternative. Other specific answers included that it was easy to use the route because there was already an existing path (9%), and that it had become a habit (11%) to use a specific route. Some of the trespassers also indicated that they had been using the route for many decades. Most people were trespassing while going shopping, jogging, or on their way to school or work. The answers differed somewhat between locations. The trespassers were also asked about the frequency of their trespassing (Table 3). Thirty-five percent of all respondents trespassed daily or almost daily. It is significant that 67% of all respondents answered that they trespassed at least once a week. This finding suggests that the majority of trespassers are those for whom trespassing is a habit and the paths across the railway lines are part of their normal routes, although they were aware of the closest official routes across the railway line.

Table 2: Frequency of Trespassing Based on Interviews.

The trespassers were also asked what they thought were the best preventive measures and what would stop them from trespassing. The respondents were allowed to indicate as many countermeasures as they wanted. First, they could spontaneously suggest different measures, after which the interviewer provided options (listed in Table 4) and the possibility to complement their answers. Both the spontaneously given suggestions and those based on the interviewer’s list were combined (Table 4).



PROCEEDINGS 85/208

Table 3: Number of Indicated Countermeasures.

The most frequently suggested measures included building a fence or an underpass/overpass. Fencing was relatively more frequently indicated at location 3 than at locations 1 and 2. The opposite applies to the building of an overpass or underpass. This difference most likely resulted from the distance to the closest official crossing site being shorter at location 3 (150 m) than at locations 1 and 2 (200–300 m). Specifically, the results suggest that people were more willing to accept fencing if the distance to the closest official crossing site was relatively short, but in the case of a relatively long distance they rather preferred an overpass or underpass. In addition to the above measures, enforcement or imposition of a fine, installation of a prohibitive sign and information provided by various means were supported. Half of the respondents assessed that the trespassing is either completely or fairly safe (Figure 2). Many of the interviewees considered trespassing safe when they are careful. Furthermore, many of them responded that they are able to cross the tracks safely and were more worried about children, elderly people, drunken people and those whose attention is somehow distracted. In contrast, about 17% of the interviewees consider trespassing very dangerous.

Figure 1: Understanding of the danger of trespassing.

Overall, 59% of the respondents considered trespassing illegal, 15% considered it legal and 26% did not know. A few respondents indicated that they had never even thought about the legality of their act. Some of the respondents also said that it must be legal, as there is no sign to indicate otherwise. In addition, at the end of the interview trespassers were able to share their views about trespassing. One of the things that emerged is that people are aware that trespassing occurs a lot in the Lappeenranta area.



PROCEEDINGS 86/208

Discussion The interviews showed that the main reason for trespassing is taking a shortcut, which confirms the results of earlier studies [e.g. 8, 11]. Specifically, the interviewees indicated that the route across the railway tracks is the shortest and the fastest alternative. Many of them have used the route for years, and according to them it is easy to use because there are already clear paths across the railway tracks. Most people were trespassing while going shopping, jogging, or on their way to school or work. Thirty five percent of all respondents trespassed daily or almost daily. It is significant that 67% of all respondents answered that they trespassed at least once a week. The most frequently suggested countermeasures along trespassers included building a fence or an underpass or overpass. In addition to the above measures, enforcement or imposition of a fine, installation of a prohibitive sign and information provided by various means were supported. The suggestions for preventive measures indicated by the trespassers were somewhat related to the locations. The measures thought to be effective at one location were not considered so effective at another. This result is supported by earlier studies that also showed that there is no single generic solution for preventing a trespass; on the contrary, a trespass tends to be specific to a location, and solutions should be tailored to specific locations and factors in order to make the implemented measures effective [e.g. 4, 11]. According to the answers in this study, the main factor that determined the suggested countermeasures was distance to the closest official crossing site. More than 17% of the trespassers considered that trespassing is very dangerous. Nevertheless they were trespassing even though the official crossing was fairly close (less than 300 m away). This result suggests that trespassers are aware of the accident risk, but are not willing to use the longer route. However, it is possible that the difference between actual behaviour and opinions was because of the interview. Presumably at least some of the interviewees wanted to appear more responsible than they actually are. Furthermore, the interview results showed that people are aware that trespassing occurs a lot in the Lappeenranta area. Many of the trespassers considered that trespassing is safe when they are careful. They indicated that they are able to cross the tracks safely but other trespassers’ behaviour may be risky. Indeed, previous research has shown that the common belief is that we are less likely to suffer negative events than our peers. This effect is called illusory invulnerability and it allows us to take risks, because the paradoxical belief is that "It will not happen to me." [5]. On the other hand, 50% of the trespassers assessed that trespassing is either fairly or highly dangerous and 15% of the respondents considered trespassing legal. Given that so many trespassers assumed that trespassing is legal, it is worth considering information campaigns to raise awareness of the illegality of trespassing and the dangers related to it as one form of preventing trespassing. Finnish Rail Agency is participating in the Operation Lifesaver, which is a non-profit, international, public education programme (originally from the United States) to prevent collisions, deaths and injuries at highway-rail grade crossings and on railroad rights-of-way [9]. Therefore, the support of Operation Lifesaver should be considered when campaigning to prevent trespassings. In addition to information campaigns, various physical measures (e.g. overpasses and underpasses, fencing, prohibitive signs) should be considered as effective and acceptable preventive measures. It is also possible to strengthen information campaigns by combining them with other measures to have more influence on trespassers' behaviour.



PROCEEDINGS 87/208

Acknowledgements Appreciation is extended to the Finnish Rail Administration for its support of the research this paper is based on. The author wishes to thank Jouni Hytönen, Mikko Kallio and Mikko Poutanen from VTT for helping with the data collection, and Juha Luoma and Veli-Pekka Kallberg from VTT for their helpful suggestions on earlier drafts of this paper. The author also wishes to thank Kirsi Pajunen from Finnish Rail Agency for all of her help during the study. References [1] Lundström, A. 2008. “Accidents to unauthorised persons and suicides”. ERA seminar on trespassers

on railway lines and suicides 3.4.2008. [2] Eurostat. Transport. European Commission. http://ec.europa.eu/eurostat/. Accessed June 6, 2007. [3] Finlex. Railway Act 555/2006. http://www.finlex.fi/en/. Accessed June 6, 2007. [4] Law, W. 2004. “Trespassing on Railway Lines - A Community Problem-Solving Guide”. CDROM. In 8th

International Level-Crossing Symposium & Managing Trespass Seminar. Rail Safety and Standards Board, Sheffield, 2004.

[5] Hatfield, J. et al. 2006. “The development of messages and experiences to reduce roadrelated illusory vulnerability and risky driving for young drivers”. Final report for the Motor Accidents Authority of NSW. http://www.irmrc.unsw.edu.au/. Accessed June 12, 2007.

[6] Lobb, B. 2006. “Trespassing on the tracks: A review of railway pedestrian safety research”. Journal of Safety Research, Vol. 37, 2006, pp. 359–365.

[7] Lobb, B. et al. 2003. “An evaluation of four types of railway pedestrian crossing safety intervention”. Accident Analysis and Prevention, Vol. 35, 2003, pp. 487–494.

[8] Lobb, B. et al. 2001. “An evaluation of a suburban railway pedestrian crossing safety programme”. Accident Analysis and Prevention, Vol. 33, 2001, pp.157–165.

[9] Operation Lifesaver. Rail safety education. http://www.oli.org/index.php. Accessed April 28, 2008. [10] Pelletier, A. 1997. “Deaths among railroad trespassers. The role of alcohol in fatal injuries”. JAMA. Vol

277. 1064–1666. [11] Rail Safety and Standards Board. “Trespass and Access via the Platform End”. Final Report. Halcrow

Group Limited in partnership with Human Engineering. http://www.rssb.co.uk/. Accessed June 12, 2007.



PROCEEDINGS 88/208


Current Status of Level Crossing Accidents and Solu tions for Enhancing LC Safety in Japan

Author(s): Shigeto Hiraguri, Kazutoshi SATO

Job Title: Research and Development

Company: Railway Technical Research Institute

Country: Japan

Resume of Speaker

In 1991, Shigeto Hiraguri joined in Railway Technical Research Institute. After training for one year, in 1992, Shigeto joined in the Train control systems laboratory of R.T.R.I. as Researcher. Engaged in a development of Computer and Radio Aided Train Control System (CARAT), a research of an efficient traffic and train control method based on information technology and a development of new train control system for secondary line have been carried out. In 2007, Shigeto was transferred to the Signalling Systems Laboratory as Head of laboratory. And in 2006, Shigeto participated in SELCAT (Safer European Level Crossing Appraisal and Technology) project. He is a Member of IRSE (Institute of Railway Signalling Engineers).

Abstract

The paper will introduce the current status of level crossings and accidents in Japan. Unique characteristics of Japanese level crossing are their density is very high, and there are many crossings in large cities. The number of level crossing accidents decreases year by year thanks to the efforts made by the government, railway operators, and road administration. However, over 400 accidents occurred last year. Therefore, it is recognised that enhancing safety at level crossings is important, and various efforts have been made. The presentation will introduce solutions that are applied currently for preventing accidents or enhancing safety at level crossings. Furthermore, we will introduce several new obstacle detecting systems which use laser radar, microwave or image processing which have been developed recently. 1 Current Status of a Level Crossing in Japan 1.1 Outline The length of all railway networks of Japan is approximately 27000 km which consists of 20000 JR networks and 7000 km private railway companies’ networks. As shown in Figure 1, the number of level crossings is 34952 which consist of 21453 level crossings of JR (former Japanese National Railway) companies and 13499 level crossings of private railway companies in 2006. The density of level crossing is very high (more than one level crossing per one kilometre) and that is a unique characteristic of Japanese level crossing. Furthermore, there are many level crossings in large cities and road traffic is generally shut off for long time because train



PROCEEDINGS 89/208

headway is short (several minutes) in such large cities. The problem is more serious if the amount of road traffic is large. In Japan, such level crossings are called “bottleneck closing” which is defined as follows. - A level crossing which is closed to road traffic at least 40 minutes in the peak hour. or - A level crossing which shuts off more than 50000 road traffic (the product of “the number of road vehicles”

and “the time of blocking road traffic”) in a day. There are approximately 1000 bottleneck crossings, and 67% of them exist in Tokyo and Osaka, the largest cities in Japan.

Figure 1 Number of level crossings in Japan

1.2 Types of Level Crossings Japanese level crossings are classified into three categories. Category-1 and Category-3 are automatically protected. Category-1 is equipped with the crossing warning sign, level crossing signals and barriers. The warning to road users is given by the crossing signal, blinking of red lights, and audio. Furthermore, the barriers protect road traffic. There are only full barrier level crossings except for minor special cases in Japan. Category-3 is equipped with the warning sign and signals but is not equipped with barriers. Category-4 is equipped with only the warning sign. In other wards, this type is not automatically protected. In 2006, there are 30188 (86.4%) category-1 level crossings, 1019 (2.9%) category-3 level crossings and 3745 (10.7%) category-4 level crossings. Category-2 is a level crossing which shut off road traffic in certain time in a day and its barrier is operated manually. However, this type does not exist because it had been upgraded to Category-1.

Table 1: Types of level crossing in Japan

1.3 Statistics of an Accident Railway operation accidents consist of train collision, train derailment, level crossing accident and so on. In all railway accidents, level crossing accidents account about 50%. Figure 2 shows the number of level crossing accidents in recent years and it decreases year by year. The reason is not only decrease of level crossing but also that various efforts to enhance safety have being made. However, 371 accidents occurred in 2006 and it is



PROCEEDINGS 90/208

equivalent of 44% of total railway accidents. Figure 3 shows the number of casualties in level crossing accidents, and the number of it decreases too, but 306 people were injured or died in 2006. A breakdown of level crossing accidents in 2006 is shown in Figure 4, Figure 5 and Figure 6. Figure 4 shows the percentages of accidents in each level crossing category. They are generally in proportion to the ratio of the number of level crossings. However, the numbers of accidents per 100 level crossings are 0.96 (Category-1), 1.57 (Category-3) and 1.71 (Category-4). It indicates that the opportunity of an accident is smaller in sophisticated level crossing. Figure 5 shows the causes of accidents. Most accidents arose from “passing a level crossing just before a train passing”. Other major causes are “automobile is forced to stop in a crossing due to engine stall and so on” and “automobile contacts or collides with side body of a train”. Contact or collision with side body occurs when a car cannot stop in front of level crossing even though a car driver brakes and so on (e.g. skids due to icy road). The percentage of these causes is 94% and it indicates that almost accidents are caused by road users. Figure 6 shows objects of collision with a train. Most objects are automobiles. On the other hand, pedestrians account about 30%.

Figure 2 Number of level crossing accidents Figure 3 Number of Casualties

Figure 4 Percentage of accidents Figure 5 Causes of LC accidents Figure 6 Objects of collision 2 Basic Facilities and Function of a Level Crossing in Japan 2.1 Facilities Figure 7 shows an example of facilities of Category-1 level crossing. It consists of a crossing warning sign which is similar to St. Andrew Cross of European level crossings, a road warning device which consist of blinking led lights with audible warning and barriers controlled automatically.



PROCEEDINGS 91/208

Figure 7 Facilities of level crossing in Japan

2.2 Function Basic control sequence of Japanese level crossing is as follows. When a train approach a level crossing, level crossing controller begins warning automatically. At first, warning sound begins and two red warning lights blink alternatively. After several seconds, a barrier shuts off road traffic. Times related to this sequence are regulated are as follows. The minimum time from the beginning of warning to the completion of shutting off is 10 seconds and the standard time is 15 seconds. The minimum time from the completion of shutting off to arriving of train at a level crossing is 15 seconds and the standard time is 20 seconds. Figure 8 shows an example of control sequence of doubled half barrier level crossing. Electronic train detectors for activating warning (A) and for stopping warning (B) are installed. When a train passes the detector A, the level crossing is activated and begins warning. After four seconds, lowering entrance side barrier starts and it is completed after six seconds. After the completion of lowering entrance side barrier, lowering exit side barrier starts and it is completed after six seconds too. In this example, the minimum warning time from the beginning of warning to the arriving of a train at the level crossing is 31 seconds.

Figure 8 Example of control sequence



PROCEEDINGS 92/208

2.3 Regulation and guidance for a Railway Operator Figure 9 shows the structure of the regulation and the guidance for railways in Japan. Facilities of a level crossing in Japan have conformity with the Regulation (Technological standard) and the Guidance (Notice) established by Ministry of Land, Infrastructure and Transport (MLIT). The Regulation is called “Shorei” in Japanese, and the Guidance is called “Kaishakukijun” in Japanese. “Shorei” specifies the target performance so as not to impede technological development, and is applied to all railway companies of any scale from companies which operate Shinkansen to minor public railway companies. An example of “Shorei” and “Kaishakukijun” is as follows. [Shorei (example)] We have to take into consideration the safety and smooth passage of a level crossing for passengers, for that purpose, the level crossing protection device shall be installed. [Kaishakukijun (example)] - The crossing warning sign shall be installed at a level crossing. - In a high speed line (over 130km/h and less 160km/h), the automatic barrier machine and the obstacle detecting device shall be installed. - A visible distance of the road warning device shall be more over 45m. Railway companies shall establish their practical standard which has conformity with “Shorei” (Regulation) and “Kaishakukijun” (Guidance).

Figure 9 Japanese ministry regulation and guidance

3 Approach to Enhance Safety of a Level Crossing 3.1 Political Measures and Legal Background The fundamental solutions to eliminate accidents are converting a level crossing into a flyover and an abolishment of it. Japanese government promotes measures such as the conversion, widening of a road, improvement of level crossing facilities and so on based on “Act on the Promotion of Improvement of Level crossing”. As the result, over 40% of intersections of road and railway have been converted into flyovers and approximately 300 level crossings are abolished every year. Following is not measures which is applied in recent years, but is unique background of Japanese level crossing. Japanese “Road Traffic Act” defines that when a car passing through a level crossing, it shall stop once in front of a level crossing and a car driver shall confirm safety. It also defines that a road user shall not go into a level crossing during warning signal is being activated, barriers being lowered or road traffic being shut off.



PROCEEDINGS 93/208

3.2 Solutions to Enhance Safety As described in 3.1, elimination of a level crossing is fundamental and most effective solution. Japanese government, railway operators and road administrator are making effort to realise it. However, there are many level crossings and it is difficult to eliminate all level crossings in a short period. Therefore, it is also important to apply measures to prevent accidents as much as possible. In other words, this approach is to secure functional safety of a level crossing. The approach is classified into following three; (1) enhancement of reliability of level crossing protection device, (2) improvement of level crossing facilities to reduce a human error of a road user and (3) introduction of devices to prevent a level crossing accident. Details of these measures are as follows. 3.2.1 Enhancement of Level Crossing Protection Devi ces’ Reliability There are two main measures to realise the enhancement of reliability. First measure is enhancement of secure train detection by introduction of an improved electronic train detector, adoption of another train detector for redundancy and modification of a train detection method from intermittently one to continuous one. Another measure is enhancement of secure device control by introduction of a micro-computerised level crossing controller. Main purpose of this type of controller is to improve non-activation or continuance of warning in case of instability behaviour of a train detecting device. Furthermore, level crossing control conditions which relates signal state in a station is complex and designing of control logic requires specific skill. Therefore, RTRI developed another type which has design assistance function. This type was introduced into many railway operators, and it contributes not only to improvement of efficiency of designing work but also to enhancement of safety. Other measures of enhancement of secure device are introduction of an observation system for quick repair in case of failure and installation of a battery as a preventive measure against failure of power supply. 3.2.2 Reduction of Road Users’ Human Error Improvement of level crossing facilities is carried out for reduction of road users’ human error. There are four main measures. First measure is to upgrade level crossing category from Category 3 or 4 to Category 1. Second measure is improvement of level crossing visibility by enhancement of maintenance of a road sign by road administrator (Figure 10), reinforcement of a lighting device for a level crossing (Figure 11), emphasis of level crossing facilities using reflective paints (Figure 12), improved barriers (Figure 13) or introduction of improved road warning devices (Figure 14).

Figure 10 Road sign installed by road administrator Figure 11 Lighting device



PROCEEDINGS 94/208

Figure 12 Emphasizing LC by reflective paint or device

Figure 13 Improved barriers

Figure 14 Improved road warning devices

Third measure is rationalisation of a warning time. The warning time of a level crossing which exists between stations is influenced by fluctuation of train speed. In this case, introduction of an improved controller using axle counters is an effective measure. On the other hand, the warning time of a level crossing which exists near a station is influence by train class (e.g. high speed train or low speed train, passing train or stopping train). In this case, introduction of improved level crossing controller using a transponder which receives information of a train class is an effective measure. Fourth measure is installation of ITV camera for vicious road users at particular level crossings. The image is not monitored continuously, but it is recorded. 3.2.3 Device to Prevent an Accident A control box to stop an approaching train in an emergency is introduced (Figure 15). Road users are able to push a button equipped with the box and he/she inform his/her operation of a railway operator. The contact telephone number of the railway operator is indicated below the box.



PROCEEDINGS 95/208

Obstruction warning devices and obstacle detecting devices are installed. There are two types of the obstruction warning device. First one is "flashing light signal" and second one is "fuse signal". An obstacle detecting device detects obstacles automatically and an obstruction warning device is activated immediately to notify an accident at a level crossing to an approaching train driver. Flashing light signal gives warning by blinking red lights. From a train driver, it seems that red right is rotating. Fuse signal gives warning by blinking red lights. A visible distance of the obstacle warning device is more than 800m, and it is longer than minimum emergency braking distance, 600m (Figure 16).

Figure 15 Control box to stop an approaching train Figure 16 Obstacle warning device There are mainly two obstacle detecting devices. Figure 17 shows the type which uses laser beam or infrared rays and it detects obstacle by shutting off of the beam. Transmitters and receivers are installed and meshes of beams are constructed in a level crossing. Figure 18 shows the type which uses loop coils and detects obstacle by change of signal frequency when a car exists above the coil. This type is used mainly in snowy area. These two types of obstacle detecting devices are used widely. However, there are problems that its detection performance is sometimes influenced by a climate condition or it requires much cost for construction and maintenance. Several new types were development to improve the problem. Figure 19 shows a new type which is installed recently. This type uses threedimensional laser radar. The radar emits a laser to an object and measures a time from emission to receiving reflected laser. The distance to an object is calculated from this time. The radar scans a measurement area in two-dimension, and measures three-dimensional shape of an obstacle. The feature of this type is that it can detect obstacle's height and detailed position in a detecting area. Figure 20 shows another new type which uses millimetre wave radars. The emission area of the millimetre wave is four degrees in vertical and 30 degrees in horizontal. The detection ratio of this type shall be seven nines for automobiles and four nines for pedestrians and so on. It is reported that this device satisfied these specifications in the field test and will be installed in practically this year. These obstacle detecting devices activate an obstacle warning device which is shown in Figure 14 or Figure 15 when an obstacle is detected during road traffic is being shut off.

Figure 17 Obstacle detecting device (1) Figure 18 Obstacle detecting device (2)



PROCEEDINGS 96/208

Figure 17 Obstacle detecting device (1) Figure 18 Obstacle detecting device (2)

Figure 20 Outline of new obstacle detecting device using millimeter wave radar

3.3 Research Project MLIT of Japan set up the research project for developing an advanced level crossing control system in 2006. The project consists of an academic experts, MLIT, National Police Agency (road administrator), and railway operators. The chairman is Prof. Takahashi, Nihon University, and Japan Transportation Planning Association is a secretariat. Main subjects of the project are investigation and analysis of the existing condition, research for a new application to reduce the interception time in level crossing and developing the more advanced level crossing system. Figure 21 shows an advanced level crossing control system which is studied in the project. Radio communication system (WRS or CRS) are located near a level crossing. This system uses special radio communication system which can detect train position without other sensors. Therefore, a level crossing control device adjusts the warning appropriately according to the velocity of approaching train. If the wireless communication is failed, level crossing control device starts the warning immediately based on conventional fail-safe control system. Other main subjects of the project are (1) establishment of quick barrier opening system using the conventional train axle sensor, (2) study on the sensing technology for train passing by application of GPS and (3) elaboration of the specific simulation engine to evaluate the interception time of level crossing facilities under given conditions.



PROCEEDINGS 97/208

Figure 21 Advanced LC control system studied in the research project by MLIT

4 Conclusions Safety of a level crossing has being important subject in Japan. Current situation and outline of a level crossing and several approaches to enhance safety of a level crossing in Japan are introduced in this paper. The number of accidents has been reduced year by year. We think that it was realised by these steady efforts. Circumstances of a level crossing and railway are deferent between countries. However, we would appreciate that solutions which are introduced in this paper could be useful reference for enhancement of level crossing safety.



PROCEEDINGS 98/208

Session 6 - Safety and Human Issues Chairman: Sue Nelson, Community Partnerships, UK

Operation Lifesaver - Reducing Level Crossing Colli sions and Trespassing Incidents through Education

Author(s): Daniel Di Tota

Job Title: National Director, Operation Lifesaver, Canada

Company: Railway Association of Canada

Country: Canada

Resume of Speaker

Dan Di Tota is the National Director of Operation Lifesaver, Canada, a public awareness program whose goals are to reduce fatalities and injuries from highway/railway crossing collisions and from trespassing on railway property. Mr. Di Tota had more than 24 years of railroading experience with Canadian Pacific Railway before being seconded to the Railway Association of Canada (RAC) in 2001 to manage the Operation Lifesaver program. Dan is fluently trilingual in English, French and Italian and was chair of the Direction 2006 Education Committee as well as a member of the program’s Executive Committee. In 2005, Dan was honoured with the “Award of Excellence” by the Transportation Association of Canada for his work and leadership within the transportation industry with the Operation Lifesaver and Direction 2006 programs. Abstract This paper will demonstrate the basic steps and considerations needed to implement an effective educational Public-Rail Safety program. It will present an overview of the Operation Lifesaver program in Canada and demonstrate the use of volunteers; their training and reporting. It will display various materials produced for a diverse audience and reveal the methods used to deliver the program using low tech methods such as flip charts and drawings as well as more advanced technology through images, computer, internet and other equipment. The presentation will also show how the program is guided by the Operation Lifesaver Advisory Board and reveal networking opportunities through vital partnerships. It includes a summary and evolution of the Operation Lifesaver Program in Canada; its strengths and weaknesses and demonstrates the successes it has achieved in helping to reduce collisions at highway/railway crossings and trespassing incidents by over 66% in the last 25 years. This presentation is intended for any organization or group contemplating the implementation of an Educational Public-Rail Safety program.

Introduction

In 1980, there were 826 collisions between trains and motor vehicles at level crossings in Canada, resulting in 83 deaths and 62 injuries. The resulting number of needless fatalities and injuries were further compounded by 177 trespassing incidents which resulted in 97 fatalities and leaving 80 more seriously injured.



PROCEEDINGS 99/208

Many of the injured were permanently disabled. Property damage from these collisions was extensive. It is no exaggeration to say that the direct and indirect costs of these incidents amounted to several million dollars annually. The greater tragedy, however, lies in the fact that virtually all of these incidents could have been avoided. Level crossing collisions are, in fact, one of the most predictable of all transportation hazards. Trains and motor vehicles alike travel on hundreds of thousands of kilometres of rail or highway and urban road networks. Similarly, aircraft have millions of kilometres of air space in which to fly. However, a level crossing has a precise location where a collision between a train and a motor vehicle is most likely to occur– the intersection of the railway track and the highway. In the years that followed, investigation reports indicated that a large number of collisions at public crossings occurred at crossings equipped with active warning devices such as gates, lights and bells; many of these collisions involved the motor vehicle striking the side of the train. These reports also revealed that, in most circumstances, motorists themselves were responsible for these collisions in that they disregarded the horn and bell warnings of approaching trains, they ignored the flashing lights and bell warnings at crossings and sometimes, they even drove around lowered barriers. Studies consistently indicated that, despite the installation of gates and other automatic warning devices, pedestrian crossovers and stricter trespassing enforcement, there was a surprising lack of knowledge about the potential hazards. These same reports, however, could not explain why people chose to trespass on railway property. Studies have shown that trespassing incidents often take place when people walk, lie down, cross or sit on the tracks. Incidents also occur when recreational vehicle users, cross-country skiers and hunters use railway property as a recreational playground. To address these issues, the railway industry joined forces with the Railway Transport Committee of the (then) Canadian Transport Commission as well as with provincial and municipal governments in ongoing and aggressive programs to improve railway-related safety. However, there was a need for a national focal point to collect information on all program efforts and to coordinate and assist all stakeholders in preventing level crossing and trespassing incidents. Therefore, to educate the public and deal with this problem, in 1981, Operation Lifesaver was established in Canada using the US model and support created in 1972.

Operation Lifesaver

Operation Lifesaver (OL) is a national public awareness program aimed at reducing railway related incidents resulting in fatalities and injuries. By using various initiatives, OL promotes safety at level crossings for drivers as well as for pedestrians. Emphasis is placed on dangerous behaviours such as trespassing on railway property or disobeying railway signs and signals. OL is a partnership between the Railway Association of Canada and its members, Transport Canada and the provincial safety councils and leagues, unions, police, public and community groups.

OL Structure

Operation Lifesaver is guided by a National Advisory Committee, which provides advice to the National Director regarding the development and implementation of the program. Members of the committee include representatives from Transport Canada, The Railway Association of Canada, Canada Safety Council, Canadian National Railway, Canadian Pacific Railway, VIA Rail Canada, GO Transit, Railway Police, Royal Canadian Mounted Police, Ontario Provincial Police, Sûreté d’Québec, First Nation’s Police Association, Teamsters Canada Rail Conference and the United Transportation Union. Provincial OL committees, made up of various stakeholders, are set-up to work on local issues and coordinate activities within their respective provinces.

Certified Presenters

The backbone of the OL Program is its volunteers. These include Master Trainers, Presenter Trainers, Presenters and Associates. Presenters and Associates are required to complete a certification process, which includes a police reference check and training. Presenters deliver the Operation Lifesaver program and Associates support the program by staffing displays and participating in other OL activities. Most volunteers come from within the rail industry. Many are railway police officers and train crews often sponsored by their organizations to give presentations and deliver the OL message. They are often first hand witnesses to the trauma and tragic events of train-vehicle or train-pedestrian collisions. Others are emergency



PROCEEDINGS 100/208

responders, teachers and individuals committed to our cause and who simply wish to help make a difference within their communities.

Why Operation Lifesaver?

In Canada, a train collides with a vehicle or a person almost every day. Many people are unfamiliar with the warning devices designed for their safety. They are unaware that trains cannot stop quickly to avoid a collision. Statistics reveal that many people are also unaware that trespassing on railway property is both dangerous and illegal. This lack of awareness emphasizes the need to educate the public on the dangers surrounding railway property. Trespassing is a leading cause of railway related fatalities. Railway tracks and bridges are private property and any unauthorized person who is on railway property, not designated as a crossing, is committing an offence under the law. Whether people are walking along the tracks, riding a snowmobile or an all-terrain-vehicle, or simply crossing a track at a location not identified as a crossing, it is trespassing. This is very dangerous and also illegal; trespassers are subject to a warning, a fine or worse, pay with their life. Operation Lifesaver is committed to working with the community and with the general public in an effort to reduce the number of fatalities and injuries due to trespassing and dangerous behaviour at crossings.

How does Operation Lifesaver meet its goals?

The very strength of Operation Lifesaver lies in its railway/government/community co-operative effort. The program seeks to join all federal, provincial and municipal authorities, enforcement agencies, railways and their unions and various public safety organizations in a nationwide effort to reduce deaths, injuries and property damages resulting from railway-related incidents. Participation by all these members is essential to the success of the program. To achieve its goals, Operation Lifesaver focuses on Education, Enforcement and Engineering – the Three E’s.

(1) Education Operation Lifesaver educates people of all ages about the potential dangers at level crossings and the seriousness of trespassing on railway property. The methods used to reach the public include the production and distribution of educational material, safety presentations to schools, industries and community groups and the staging of media events.

(2) Enforcement

Operation Lifesaver promotes enforcement of laws governing motorist and pedestrian responsibilities at level crossings and on railway property. The program works in partnership with enforcement agencies to heighten public awareness of the hazards which exist with respect to railway property.

(3) Engineering

Operation Lifesaver promotes research aimed at ensuring a high level of safety at railway crossings and on railway property. Operation Lifesaver also informs the public of federal, provincial and other programs aimed at improving railway safety.

By focusing on these issues and identifying emergent concerns, Operation Lifesaver has been able to help reduce level crossing collisions by more than 75% and trespassing incidents by over 44%.

Presentations/Activities

Operation Lifesaver has special audio-visual programs and printed material available to help. The key to the success of Operation Lifesaver is through active participation which can be simply familiarizing the public with the appropriate warning signs.



PROCEEDINGS 101/208

The OL program is designed to increase general public awareness of potential hazards at level crossings and of the dangers associated with trespassing on railway property. This program can be adapted to any age group and is given in the form of a presentation. Presentations are scripted and provided to all our certified presenters and can last between 30 and 60 minutes, depending on the audience and its needs and is offered to schools and organizations using various methods. They include:

(1) Basic Visual Presentation Using basic diagrams on flip charts or poster boards, the presenter demonstrates common hazards at level crossings and surrounding railway property.

(2) Audio-visual Presentation Presentations can be supplemented with the use of audio-visual equipment featuring Operation Lifesaver produced film clips and slides.



PROCEEDINGS 102/208

(3) Mock simulation presentation/activity Through visual stimulation of a simulated crash, audiences can experience the trauma and effects of a train/vehicle collision and witness the efforts of emergency responders as they manage the scene.

(4) Display booth activity At fairs and events, display booths allow opportunities for our presenters and associates to deliver OL material such as pamphlets, brochures, CD’s, activity books and more.

(5) Crossing blitz activity Law enforcement partners pass on life saving tips and brochures to motorists and pedestrians as they approach level crossings during this carefully planned initiative.



PROCEEDINGS 103/208

(6) Officer on a train activity This initiative involves several law enforcement officers shadowing a train on which they have a colleague monitoring driver behaviour. Illegal and dangerous actions are reported back to the following cruisers. Violators are stopped, informed of their dangerous behaviour and educated on the risks they place on themselves and others.

(7) Platform blitz Similar to a crossing blitz, presenters circulate on commuter platforms distributing OL material and highlighting the dangers of being too close to an approaching train.

(8) Media Activity This activity may entail several variations. Prepared press releases, public service announcements and radio or television interviews are to name a few.



PROCEEDINGS 104/208

Operation Lifesaver also offers several stand-alone information packages for Professional Drivers and Newly Licensed Drivers to learn about safety around Railway property. These packages are provided free of charge to all Industry Instructors and are easily incorporated into their existing curriculum. These packages are available for Truck, School Bus, Emergency Responders, Motor Coach and Transit drivers. In order to help deliver our message to young drivers, OL has also developed and interactive website to teach new drivers of the signs and hazards approaching level crossings. It can be located at www.traintodrive.net.

Conclusion

In 2007, the Operation Lifesaver, Canada program had 360 active presenters participate in over 1800 different presentations and activities while reaching out to more than 1.7 million Canadians across the country. The Operation Lifesaver program owes a great deal of gratitude to the continued contribution from its partners, stakeholders and volunteers without whom; the successes we have achieved would not be possible. Their commitment and determination to prevent another tragic event inspires us to continue in our efforts to produce quality and appropriate safety materials and schedule presentations to a variety of audiences across the country. [1] Based on the OL “Look, Listen and Live” brochure with input from the Operation Lifesaver Program Review Committee (OL PRC), Canada.



PROCEEDINGS 105/208


Rail Safety Education Strategies in the United Stat es

Author(s): Helen Sramek

Job Title: President

Company: Operation Lifesaver, Inc.

Country: USA

Resume of Speaker

Helen Sramek became President of Operation Lifesaver, Inc. in January, 2007. She comes to the organisation from AAA where she served for 9 years as Director of Federal Relations. In that capacity, she led AAA’s advocacy efforts on federal transportation, traffic safety and consumer issues before Congress and the Administration. Sramek, a native Nebraskan, has worked in a variety of positions in Washington, D.C. for over 30 years. In addition to considerable legislative experience, Sramek served as Chief of Staff to a Nebraska Congressman for over a decade. She played a key role in helping the Congressman secure federal support for a major highway-rail grade crossing in the state’s capital city of Lincoln, Nebraska. She also brings federal Executive Branch experience to her role at OLI, having served in the first Bush Administration. Operation Lifesaver is a non-profit, international continuing public education program first established in 1972 to end tragic crashes, fatalities and injuries at highway-rail intersections and on railroad rights-of-way.

Abstract This paper will demonstrate tactics and strategies for expanding outreach to the government, private sector companies, and other safety groups in order to leverage the effectiveness and resources dedicated to a public rail safety education program. It will present background on and a description of the Operation Lifesaver (OL) program in the United States, and its proven model for reducing collisions, injuries and fatalities at level crossings. The presentation will also show how Operation Lifesaver is analyzing its safety education efforts for pedestrians, with the goal of refocusing its pedestrian rail safety education materials and general presentations to target the growing U.S. pedestrian rail trespass problem. This effort is a work in progress that is changing the way OL trainers and presenters approach the trespass issue. Introduction The Operation Lifesaver program began over 35 years ago as a six-week public relations campaign launched in the state of Idaho by Union Pacific Railroad in 1972. Word of its success spread to other states which initiated similar programs. Between 1978 and 1986, while Operation Lifesaver operated under the auspices of the National Safety Council, all 49 continental states started independent Operation Lifesaver programs. In



PROCEEDINGS 106/208

1986, the national program became independent and incorporated as a national, non-profit, 501(c)(3) educational organization. Today Operation Lifesaver, Inc. (OLI) stands as a non-profit national support center for the nation’s premier highway-rail safety education organization. Its strength resides with its state programs—all of which are independently organized, yet serve as members of the national organization to ensure consistency of message and access to the latest tools for delivering the OL message. OL’s mission is clearly stated: to end collisions, deaths and injuries at highway-rail grade crossings and on railroad rights-of way. That first Operation Lifesaver program in Idaho consisted of a slide presentation, accompanied by an audio message that lasted fifteen minutes. Today, Operation Lifesaver reaches millions of people through presentations that can be customized with safety information for different audiences. A variety of communications techniques are utilized ranging from the traditional “flip charts” to power point presentations and specialized training videos. Through the OLI website, users can download public service announcements, customized artwork, and lesson plans for student groups and teachers. In April, 2008 OLI released its newest training video for driver education students. The interactive video, presented by secondary school students in a peer-to-peer format, includes a web-based, e-learning component that allows many more new drivers to have access to the latest rail safety information. In 2007, Operation Lifesaver and its state programs reached 3.7 million people with its life-saving message.3 Several million more were reached through the OLI website.4 A program that began small has evolved into a nationwide organization of trained volunteers whose commitment to safety launched a movement that continues to grow. Operation Lifesaver is a success story within the U.S. traffic safety community. Vehicle-train collisions have dropped by more than 75 percent nationwide since 1972, the beginning year for Operation Lifesaver.5 A variety of safety efforts are credited with this drop, particularly an ongoing safety partnership among federal, state and local governments, law enforcement agencies, traffic safety organizations, railroad companies and OL. A concerted effort to close or eliminate level crossings has also contributed to these safety gains. This improvement comes despite increases in vehicles on the road, trains on the rails, and the expansion of light rail and commuter rail lines. By combining the activities of engineering, enforcement and education, Operation Lifesaver has contributed to a traffic safety success story. In the last 20 years alone, crossing crashes have been reduced by fifty percent.6 Preliminary U.S. government statistics for 2007 show there were 2,741 crossing incidents, down from 6,617 in 1988; 339 fatalities from vehicle-train collisions, vs. 689 in 1988; and 1,003 injuries from crossing collisions, vs. 2,589 in 1988.7 When it comes to rail safety, Operation Lifesaver is working to bring those numbers to zero. Other countries have adopted Operation Lifesaver’s public rail safety education model. In 1981, an Operation Lifesaver program was started in Canada.8 The United Kingdom’s Track Off campaign is managed by the Rail

3 Source: 2007 Operation Lifesaver annual reports.

4 Source: Website monitoring reports.

5 Source: Federal Railroad Administration Office of Safety statistics, available at

http://safetydata.fra.dot.gov/OfficeofSafety/.

6 Ibid.

7 Ibid.

8 Website: http://www.operationlifesaver.ca/.



PROCEEDINGS 107/208

Safety and Standards Board.9 Estonia has started its own program, Operation Lifesaver Estonia (OLE).10 Argentina and Mexico have also worked on developing rail safety education programs. Partnerships Key to Operation Lifesaver Success Those who led efforts to build Operation Lifesaver into the national, indeed international, organization it is today knew from the outset that building partnerships with stakeholder groups was essential in leveraging the resources of a small non-profit organization. That philosophy is embedded into the bylaws of the organization and its governing structure. The three founding partners - the Association of American Railroads, the Railway Progress Institute (now the Railway Supply Institute) and the National Railroad Passenger Corporation (Amtrak) - hold permanent positions on the OLI Board of Directors. The Board itself is broad-based and includes representatives from the organizations representing state public transportation departments, law enforcement, public transportation, as well as a state coordinator. It was also recognized very early in OLI’s development that there was a need for a representative group of partners to help shape programs and development of educational materials. Thus, the Program Development Council was created. Its mission is to bring consensus to the development and implementation of all Operation Lifesaver programs and materials. The PDC is an advisory body to the national office and includes a wide range of partners with representatives from railroads, professional drivers, state and federal public agencies, the engineering and enforcement communities, and other organizations. Two positions on the OLI Board are reserved for the Chair and Chair-elect of the PDC. Operation Lifesaver’s success is in no small measure attributed to its embracing a wide range of partners to accomplish its purposes. Its structure, funding, and ability to conduct successful public outreach programs all hinge on its established partnerships. Over the years, dozens of organizations have partnered with OLI to both develop and maximize the group’s public education efforts. Two recent examples of safety partnerships are worth noting.

• One of the most successful programs that OLI has conducted has been its partnership with law enforcement. When a collision occurs at a highway-rail intersection, law enforcement officers respond. Familiarity with railroad operations and the conditions at highway-rail grade crossings is important for the safety of law enforcement officers. Operation Lifesaver’s Grade Crossing Collision Investigation (GCCI) course provides safety training to local law enforcement officials about how to remain safe when investigating a train collision. This training was developed for the North American law enforcement community with the cooperation of the International Association of Chiefs of Police and the National Sheriffs Association. The GCCI curriculum is being revised, and OLI will be seeking copyright protection for the course. To enhance the value of the curriculum, OLI will be partnering with the International Association of Chiefs of Police, which has agreed to co-copyright the course.

• In April, OLI unveiled its new training video on rail safety for novice drivers, Look to Live. In addition to

making the video available to Operation Lifesaver presenters, OLI is working with leading driver training organizations like the American Driver Safety Education Association, the Driving School Association of the Americas, Inc. and AAA, the American nonprofit automotive service organization. This approach will enable Operation Lifesaver to reach many more new drivers with information about practicing safe driving behavior at level grade crossings. Consider the fact that in 2007 Operation Lifesaver reached almost 200,000 novice drivers with presentations. 11 That number is impressive until one realizes that there were over 3 million novice drivers in 2007 who registered for some type of formal drivers’ education class12. The new training video enables Operation Lifesaver to reach many more young drivers with life-saving tips about practicing safe driving behavior on or near railroad tracks.

9 Website: http://www.trackoff.org/community.htm.

10 Website: http://www.ole.ee/.

11 Source: 2007 Operation Lifesaver annual reports.

12 American Driver Safety Education Association, http://www.adtsea.iup.edu/adtsea/.



PROCEEDINGS 108/208

State Programs Thrive on Partnerships While partnerships are important to the national Operation Lifesaver program, they are the lifeline of state programs. In many cases, individual OL state programs are small, often managed by part-time coordinators. They have no choice but to actively seek out partners who help them expand their educational outreach throughout the state. Success stories in one state are shared through the national office and become “best practices” that other state programs can replicate. An excellent example is the partnership that some states have pursued with professional sports teams. The first state program to pursue a partnership was Georgia Operation Lifesaver, which initiated ongoing Operation Lifesaver publicity events with the Atlanta Braves baseball team. Other state programs have developed similar partnerships with their sports teams. Last year, North Carolina Operation Lifesaver partnered with other organizations in the state to release a television public service announcement that featured National Association for Stock Car Auto Racing (NASCAR) race car driver Casey Mears. Many state Operation Lifesaver programs partner with local law enforcement officials to raise public awareness of the rules and regulations that apply to grade crossing safety. State OL programs work with law enforcement agencies and railroads to run “Officer on the Train” events along selected rail corridors to raise awareness of the crossing safety issue. The train’s locomotive is fitted with cameras to capture activity at the crossing while the train is en route; monitors are stationed in passenger cars on the train so that local media, government and law enforcement officials, who are invited to ride the train, can see what happens on a real-time basis. In addition, law enforcement officers are stationed at points along the train’s route to write tickets for motorists who do not obey the crossing signs and signals. Another law enforcement-related activity used by state OL programs to highlight the crossing issue is a crossing “blitz.” OL representatives and law enforcement officials go to high-risk highway-rail intersections for a day or two and distribute safety information to motorists, reinforcing the need for caution with motorists in the community. News media are notified of these types of community activities, and the resulting television, radio and print news coverage is crucial to delivering OL safety messages to a broader audience. Light Rail and Transit Agency Efforts For several years, Operation Lifesaver has also been lending its safety expertise to public transit agencies around the country. As more metropolitan areas turn to public transportation as a means to reduce congestion, the exposure for increased incidents between people, vehicles, and commuter or light rail trains increases. In partnership with the Federal Transit Administration, OLI has produced light rail safety information materials in a variety of formats that transit agencies are using to educate their riders about rail safety. Specifically, Operation Lifesaver has worked on pilot partnership programs in Phoenix, Arizona; Southern California, Salt Lake City, Utah, Seattle, Washington, and Charlotte, North Carolina. Transit agencies recognize that Operation Lifesaver’s education efforts, which have worked so well for freight rail safety, may help prevent incidents, deaths and injuries involving public transportation. OL will continue its educational partnerships with established transit systems, including Houston, Texas, San Francisco, California and, St. Louis, Missouri. Keys to Effective Partnerships A successful rail safety education program utilizing the Operation Lifesaver model requires establishing solid partnerships from inception through implementation. The U.S. program has also benefited because it began at the grassroots level, which fostered a committed group of volunteers to spread OL’s lifesaving message. To summarize:

• A rail safety education campaign requires the support of a wide variety of community stakeholders to effectively disseminate its messages. Education safety campaigns are unlikely to have sufficient funding to reach the masses without the additional support of key stakeholders.



PROCEEDINGS 109/208

• Reducing collisions, fatalities and injuries at highway-rail crossings should never be viewed only as a railroad issue; it’s a highway safety issue requiring a concerted effort from the broader traffic safety community.

• Seek opportunities to incorporate crossing safety concerns at the local level, through existing organizations. An example in the U.S. are local metropolitan planning organizations (MPO's), which, under the 2005 federal law that provides funding for U.S. highways, highway safety and transit, must develop transportation plans and programs for metropolitan areas.13 Through participation in the MPO process, Operation Lifesaver state coordinators have been developing new relationships and finding opportunities for safety partnerships.

• Establishing and maintaining ties to key stakeholders is an on-going priority. Once a partnership is established to advance a particular project, it’s important to build and strengthen the partnership on an on-going basis.

Pedestrian Safety - an Increasing Challenge Railroad safety advocates in the United States are grappling with an issue of growing concern: the number of pedestrians killed while walking on or near railroad tracks has increased, while deaths in level crossing incidents have dropped. Operation Lifesaver is increasing its focus on trespass prevention efforts, while continuing its strong efforts in the area of level crossing safety. In the United States, railroad tracks are private property. Pedestrians who choose to walk on train tracks or gather near tracks are trespassing on private property. They are breaking the law. Since 1997, annual pedestrian/trespass fatalities in the U.S. have surpassed grade crossing fatalities.14 In 1987, 453 deaths resulted from trespassing incidents.15 Last year’s preliminary fatality numbers show 486 trespass fatalities – an increase of 7% from 20 years ago.16 U.S. railroads and rail safety advocates are renewing efforts to address the trespass problem, and Operation Lifesaver is very much a part of these efforts. The OL message, “Stay Off, Stay Away, Stay Alive” is targeted to potential rail trespassers. Addressing the trespass issue, however, is complex. People who engage in this risky behavior do so for a variety of reasons, making it difficult to customize a public information campaign. A March, 2008 Federal Railroad Administration (FRA) study sought to describe the typical trespasser, but the study’s data are limited, making it hard to generalize results across the U.S. In addition, many pedestrian rail trespass fatalities involve alcohol and drugs, societal issues that Operation Lifesaver’s programs are not intended to address. The FRA report, Rail-Trespasser Fatalities: Developing Demographic Profiles, 17 strongly recommends additional demographic analysis to develop targeted educational and outreach programs and law enforcement initiatives aimed at preventing rail trespassing incidents.

13

Source: Statewide Transportation Planning; Metropolitan Transportation Planning; Final Rule (effective 3/16/07);

Federal Register, February 14, 2007.

14 Source: Federal Railroad Administration, Office of Safety, Op.cit.

15 Ibid.

16 Ibid.

17 George, Bruce, Rail-Trespasser Fatalities: Developing Demographic Profiles, Federal Railroad Administration, March

2008, available at http://www.fra.dot.gov/downloads/safety/tdreport_final.pdf



PROCEEDINGS 110/208

As pedestrians, Americans tend to be risk-taking and impatient, believing they have the right-of-way whether standing on a street corner ready to step into an intersection or walking across what appears to be a deserted stretch of railroad tracks. They are complacent when it comes to their safety around railroad tracks. The problem is exacerbated by media images that romanticize foolish, dangerous behavior on or around railroad tracks. Train tracks are seen as a great backdrop for photo shoots – not as a source of potential danger that could lead to the loss of life. A study from Finland, presented at the January 2008 annual Transportation Research Board meeting in Washington, DC18, gives weight to the fact that pedestrians do not know that it’s illegal or dangerous to be on railroad tracks or property. The researchers interviewed trespassers crossing at sites within 500 meters or less of authorized crossings. Nearly a third did not know their actions were illegal, and 15 percent thought it was legal to be on the tracks.19 OLI’s Trespass Prevention Initiative In 2007, to respond to the growing issue of pedestrian rail trespass incidents, Operation Lifesaver, Inc. initiated a broader rail trespass prevention effort with an in-depth assessment of its trespass prevention information and programs. As a result of this assessment, Operation Lifesaver identified the following priorities for its ongoing trespass prevention program:

• Ensure that the trespass message is part of every OL presentation; • Identify new venues for OL’s educational outreach, such as soup kitchens, homeless shelters, and

college hangouts; • Find new ways to educate risk-taking youth; • Work with the engineering community to consider developing signage with easily-understandable

graphic images instead of text, for non-English speaking populations; • Partner with local law enforcement agencies on community trespassing education blitzes; • Engage community leaders to make them integral to the outreach efforts; • Continue to support and expand transit and light rail education efforts; and • Partner with the U.S. Department of Homeland Security to link OL’s safety message with national

security initiatives. Future Efforts

Operation Lifesaver will continue its two-pronged efforts to reduce rail-related incidents, deaths and injuries, working with its safety partners in the public and private sectors. These partnerships have proven to be an effective method of disseminating rail safety education materials and messages to key audiences in the United States. This model is increasingly being used around the world.

18

“Trespassing on Finnish Railways: Identification of Problem Sites and Description of Trespassing Behavior” by Anne

Riikka Silla and Juha Luoma, VTT Technical Research Centre of Finland, presented at Transportation Research Board

Annual Meeting. Session 557, January 15, 2008.

19 Ibid.



PROCEEDINGS 111/208


A new educational approach for preventing Level Cro ssing Accidents

Author(s): Subhasis Ganguly

Job Title: Chief Safety Officer

Company: South Eastern Railway (Indian Railways)

Country: India

Resume of Speaker

Subhasis Ganguly, born in 1955 joined the Indian Railways Service of Signal Engineers in 1978 after graduating in Physics in 1973 and in Electronics Engineering in 1977 from Indian Institute of Technology, Madras. He served in all most all the important posts in Signalling and Telecommunications Engineering in various capacities over at least seven different Zones of Indian Railways up to 2005. He has also worked in the Research, Design & Standards Organisation, Lucknow as Director, both in quality control & standardisation. He has been trained in Germany, USA, UK and Canada in ‘Stored Programme Control Exchanges, High Speed Signalling and Electronic Interlocking. His interest in Railway safety matters kindled while working as a Signal Engineer, finally brought him to work as the Chief Safety Officer of the South Eastern Railway in June,2005,in which capacity, he is still continuing.

Abstract

In the Indian Railways there are 16550 manned level crossing gates and 21800 unmanned level crossings according to data of July 2007. The manned level crossings gates are manned round the clock in shifts of durations governed by the density of the rail traffic, where as the unmanned level crossings are usually on rural unpaved roads intersecting block sections where both the road density and the rail traffic density are small. However, the numbers of accidents in the unmanned level crossings are much higher averaging to approximately 80 consequential accidents per year. The trend of accidents at unmanned level crossing gates appears to be increasing. It does not look possible to have any cost effective and maintainable method to warn and protect road vehicles in the unmanned level crossings over such a large network. The technical solutions adopted can be put under the following heads: Elimination of all such level crossings: Conventional wisdom suggests putting gate barriers and manning them involving both initial as well as recurring expenditure. Road traffic at many such locations is low, and seasonal. Construction of normal grade separators is quite expensive. Provision of cost-effective limited-use underpasses is already taken up. However this is possible only on high banks and therefore is of limited use. Train approach-warning systems: These are under trial using various technologies. A reliable, tamper-proof, cost effective system is some way off. In the electrically noisy, temperature wise extreme, country where some digging is always going on such systems do pose reliability and maintainability issues. Absence of service roads and accessibility of trackside devices in block sections are also issues. Evidences suggest that motorists



PROCEEDINGS 112/208

involved in many of the level crossing accidents were aware of trains approaching, yet tried to cross the track taking a calculated risk out of bravado or urgency. How effective will approach-warning devices be, is therefore a matter, which still requires to be conclusively established. Pending elimination of level crossings in the long run, Indian Railways are attempting to reduce the risks at such crossings, by conducting sustained public awareness campaigns The Indian Railways have adopted different strategies to communicate with the road users. They are also trying to understand the human motivational factors behind the high-risk behaviour of the road users at the level crossings. This is a multi pronged approach and uses print media, TV and electronic media, massive SMS campaigns, advertisements, road shows and even using celebrities to rivet the interest of the Public. The provisions of the Indian Motor Vehicle act are quite stiff, and wherever possible cases are booked against the errant road users if they survive the accident. It is to be remembered however that the heavy population densities in the country, higher aspiration levels and robust economic growth are all contributing towards increasing road congestions and long rows of waiting vehicles at level crossing gates. The migration from foot to the bicycle, from bicycle to the motor cycles, from motor cycles to cars, from economy cars to larger cars and correspondingly from light commercial vehicles to heavy and very heavy commercial vehicles are all adding up to congestion at least for short periods in village roads also. Coupled with a general sense of hurry, impatience, bravado and propensity of the road users to take risks, the numbers of level crossing accidents do not seem to respond to the preventive measures, as much as they should. The state governments, the municipalities and the village communities called ‘Panchayats’ are also being roped in. The idea is to change the perception of the public of level crossing accidents from railway accidents to road accidents, which have essentially happened because the railway’s right of way has been trespassed. The objective of this paper is to clearly set out the issues of working in partnership with of local communities which is considered very important, the achievements and shortcomings in understanding the road users’ behaviour and suggest solutions to prevent trespass being done. Apart from enforcing the legal provisions of the Indian Railways’ Act and Motor Vehicle Act, it is believed that most of the problems of risk management in level crossings can be reduced in a cost effective manner by going in for campaigns, and attacking the soft issues so that the more costly alternatives like manning, interlocking constructing over bridges or installation and maintenance of train activated device can be kept to the minimum. By doing this the totally unnecessary loss of life, the operational disruption resulting out of level crossing accidents and the legal complications can be reduced. This would result in lesser capital cost, lesser cost of operation and better safety record of Indian Railways, reassuring the Indian public that the IR is doing all it can not only to increase the safety of its passengers, but also to the road users who have an equal contribution to make for ensuring their own safety at level crossings. Introduction In the Indian Railways there are 16550 manned level crossing gates and 21800 unmanned level crossings according to data of July 2007. The manned level crossings gates are manned round the clock in shifts of durations governed by the density of the rail traffic, where as the unmanned level crossings are usually on rural unpaved roads intersecting block sections where both the road density and the rail traffic density are small. However, the numbers of accidents in the unmanned level crossings are much higher averaging to approximately 80 consequential accidents per year. The trend of accidents at unmanned level crossing gates appears to be increasing. It does not look possible to have any cost effective and maintainable method to warn and protect road vehicles in the unmanned level crossings over such a large network. Accidents at unmanned crossings constitute 80% of the total consequential accidents. Persons killed in Level Crossing Accidents constitute almost 90% of the total persons killed in train accidents. Accidents at unmanned level crossings constitute almost 90% of the total accidents occurring at level crossings. The Indian Railways rules provide for five classes of level crossings working over, four categories of roads. The five classes of level crossings are (a) Special class, (b) ‘A’ class (c) ‘B’ class (d) ‘C’ class meant for roads and (e) ‘D’ Class only for cattle. These classes are settled on the basis of train vehicle units (TVU). This is a simple multiplication of the number of trains going in both directions into the number of powered wheeled vehicles with wheels more than two. It is obvious that motorcycles, scooters, bi-cycles and bullock carts are not included in this. Any level crossing above five thousand TVU’s qualifies for manning .No unmanned level crossings are allowed within station limits and gates more than twenty five thousand TVU s automatically qualify for



PROCEEDINGS 113/208

interlocking with train signals so that unless the gates are closed the departure signals for the trains cannot be cleared. It stands to reason that all gates within suburban sections are interlocked. The trains in all automatic signalling territories, which by and large coincide with the suburban sections, back lock the barrier operating mechanisms This is further compounded by having two categories of gates one in which the gates are normally opened to road traffic and the trains can only pass if the gates are closed. The other in which the gates are normally closed to road traffic and the trains have the normal right of way and the gates are opened only to allow waiting road vehicle to pass if there is no train. The gates are normally supposed to be kept open to road traffic if the TVU is above Fifteen Thousand. The problem is that because of heavy population and even larger growth in rail traffic to the tune of almost 10% every year on my railway even class two roads and class three earth roads easily exceed the Fifteen Thousand mark. Even if we maintain a reasonably good surface in the level crossing area the flow from these roads gets clogged up and usually vehicles jostle with each other for right of way on the good and hard surface of the level crossing where a free for all results. This leads to detention of trains upsetting of schedules and eventual resentment of customers. The other solution of keeping the gates closed to road traffic either interlocked or non-interlocked doesn’t help as the departure signals of a station are only taken off when the gate is reported to be closed by the gateman. The gates across a Twelve Kilometre section may remain closed for even half an hour depending upon the speed restriction of force and the status of the train this leads to resentment among the road users where the queues as long as One Kilometre in both directions. The gateman is a lone man doing sometimes ten hours duty and falls an easy prey to assault by truckers assisted by the local tuffs and a handful of feudally minded people who feel that their stature in the village is reduced if they are stopped at a gate. Depending upon patterns, police protection is provided but over such a large railway regular police protection is impossible and usually ineffective. The incidents of assaults on gateman man leads to managerial problems in a highly Unionised work force Manning and interlocking of barriers is further constrained by the fact that in so called special class gates where the roads can be Nine Meters or more wide in busy sections the TVU’s can be as high as Four Hundred Thousand or more. This amounts to Four Thousand vehicles in a day in a very busy section where hundred trains are passing in a day. This is equivalent to a road vehicle every twenty seconds or so. It may be well understood that neither the train nor the road vehicle will ever be free of detention. There are hundreds of such gates across the Indian Railways and each gate is a potential safety hazards. Though Rumble strips and speed breakers are provided there are known instances of the speed breakers having been removed or damaged to facilitate movements of road vehicles. On my railway there are instances of at least Five gates which are deliberately broken by the truckers so that inter-locking fails and the roads have an undisputed way across the tracks as all trains follow emergency procedures to stop and look before crossing the gate. Half barriers also do not work, because the speed mixes in Indian roads as well as on rails vary widely. As a result the anticipation of the road user, as well as that of the train driver were found to be mostly wrong and therefore the half barrier approach, though tried was given up rather quickly. To solve all these issues the first thing that comes to mind is grade separation. The national highways authority has been now doing this as routine in all four lanes or higher high ways .The work on the golden quadrilateral connecting the four major metros is almost complete and this has resulted in reduction of a substantial no of very busy level crossings. However the focus of the problem has shifted elsewhere. The pace of urbanisation in India is rather high, most of the new urban centres have grown out of erstwhile large villages, and small towns, which had population of a few hundred thousands only a few years ago. Needless to say that such growth has been mostly haphazard, unregulated and has usually been across the two sides of roads and railway track. What was originally only a mud track for cattle has suddenly become a level crossing. The traffic across such roads are sporadic and happen in clusters .For instance a new small quarry opening up in the hillocks adjacent to the track, means sudden increase in traffic of dumpers and earthmoving machinery. New hutments spring up over night and the crossing becomes busy. By the time the railways take the higher TVU into account in their biennial exercise and manage the infrastructure work and recruitment of persons for manning the gates is gone through the crossing has already taken its toll. The Railways in India are state owned and the necessity to control Government expenditure on staff, which is a recurring cost, is an issue, which results in protracted correspondence with the Finance wing. As a village metamorphoses in a city the traffic mix also changes, in a matter of a few years the traffic of pedestrians and bicyclists gives way to unwieldy tractors with trailers, converted jeeps, auto rickshaws the, light commercial vehicles, and Trucks usually carrying building materials and food stuff. There are rickety contraptions built on rickshaws by attaching motorcycle engines, a V belt drive



PROCEEDINGS 114/208

with a tensioning device as clutch and plain hand lever operated bicycle brakes. Indians as a people are very innovative and skilled and many of these vehicles are not defined in any automobile dictionary. These are purchased out of bank loans and used ruthlessly as a means of livelihood till they are complete junk. The ever increasing cost of motor fuels, and a competitive urge to keep the fares, down ensures that all these vehicles are overloaded to the gills, for short to medium rides and carriage of goods between villages. This ensures that comfort and reliability are not important; you can always take the next one, if this one breaks down. Unfortunately because of the territory most of our railway track is in low to high embankments or in shallow to deep cuttings. The ‘B’ class and ‘C’ class gates allow approach roads with 1in 30 and 1in 20gradients respectively on approach roads to level crossings. The ‘A’ class gates are on gentle approach slopes of 1in 40,are subject to more aware road users of the cities and are negotiated by fitter road vehicles. The ‘C’ class level crossings across narrower roads therefore pose the biggest safety hazards, contributing to 90% of the accidents. This is because most of the vehicles of descriptions given above barely manage to negotiate the approach gradients and fail on the tracks. Almost all of our accidents happen this way. This then is the Indian scenario probably very different from the scenario in Europe. Train approach-warning systems are under trial using various technologies ranging right from track circuits and axle counter based devices to GPS and secure radio-based systems. A reliable, tamper-proof, cost effective system does not seem to appear at the end of the tunnel. Indian power supply conditions are very noisy and result in disturbances to digital axle counter signals causing them to fail. The temperature in the trackside boxes quite often goes beyond 60 Celsius. This affects the life and reliability of trackside transducers. Track circuit devices also fail because track side digging for agriculture, forestation purposes and various Railway or civil construction activity always goes on forever increasing in intensity because every busy level crossing signifies the growth of the surrounding locality as brought out earlier. Absence of service roads and accessibility of trackside devices in block sections are also issues. Every failure is a safe side failure in the railways’ sense. As far as the road user is concerned it is the hooters or flasher going off on their own, without any perceptible reason. He therefore treats the warnings with contempt and scorn. Almost all our accidents have revealed that motorists involved in the level crossing accidents were aware of trains approaching, yet tried to cross the track. Last year there were two incidents where bystanders had orally warned the motorists of approaching trains, but their advice was ignored. Taking a calculated risk out of bravado or urgency is considered macho. How effective will approach-warning devices therefore be, remains a matter of concern. I have dealt with the problem of putting gate barriers and manning them involving both initial as well as recurring expenditure in the earlier paragraphs of the paper. At the present juncture this solution suggested by conventional wisdom does not appear justified on both initial and lifecycle costs. Road traffic at many such locations is low, and seasonal. Construction of normal grade separators is quite capital intensive and involves complicated cost sharing methods between the central and state governments. Provision of cost-effective limited use underpasses is already taken up using IR’s own finances. However this is possible only on high banks and therefore is of less than adequate benefit. The closure of already existing gates is a matter where the local administration has to agree in consultation with the local population and then the permission of Commissioner of Railway Safety has to be taken. Needless to say that in a democratic country getting this sort of an agreement from the local people involves many rounds of protracted negotiations. Even then on the South Eastern railway we had succeeded in closing 22 level crossings. Reduction of risks by merging of roads and level crossings is another controversial issue. No individual or organised group readily agrees to merge two level crossing gates say a kilometre apart, and thereby being forced to take a detour of two kilometres for the perceived risk of having a level crossing accident. As far as they are concerned it is always too distant an event and could not happen to them. The IR is worried also about the visibility of an approaching train to the road user. This is considered adequate if from a distance of 4.5 metres from the centre of the track, a road user can see 600 metres of the track clearly. This is the minimum, if the visibility is less than this; speed restrictions are put depending upon the severity of the visibility shortcomings. The IR has fixed up certain minimum standards and specifications for the level crossings. The IR broadly becomes responsible for some exgratia payments and compensation payments in case the level crossing accident is attributable to some shortcoming of the IR. However the reverse does not apply. Though the Indian Railway Act and the Indian Motor Vehicle Act cover level crossing accidents and prescribe heavy penalties, the violator of the level crossings the road user pays for his negligence with his life and therefore the stringent provision of the laws end up being nothing more than good intentions. As far as



PROCEEDINGS 115/208

catching those who take risks at level crossings and escape is concerned in a given date there are millions of them spread over Sixty Thousand kilometres and the very size of the problem prevents the law being enforced quickly It is therefore considered that a new educational and mass contact approach to the level crossing safety problems is likely to yield better safety performance than more and more technological inputs and legal hassles. The problem in the Indian context particularly requires to be looked into from awareness and voluntary compliance angle rather than a purely technical and legal angle. The corner stone of such an educative campaign is to bring home the fact that the loss of life and property in level crossing accidents can be totally eliminated if only the right of way of the railway over the railway track is recognised and certain very elementary precautions are taken by the road users. This appears to be a better solution than any complicated technical schemes whose efficacy is at best doubtful because of the social technical and financial framework in a society, which is undergoing a fast transformation. Such an approach was based on the following premises.

1. A mass contact campaign should avoid emphasising on the coercive element of emphatically pointing out the penalties provided for in the statutes. In a heavily populated and thinly policed, country where the police is even now identified as a feudal, colonial and corrupt force, in the eyes of the common man, invoking the possibilities of imprisonment at best provokes mirth and disdain, and at worse provokes anger and an attitude of let us see. The response that this will not happen to me is also fairly common. The concept of social peace and harmony in India has always been a matter of long negotiation between protagonists through voluntary reconciliation processes. The attempts at mass education hitherto taken by the various railways under the IR were aware of this fact, but chose to turn a blind eye because of inbred conviction that the government knows best. With the opening up of the economy the FMCG sector realised that the untapped markets in the hinterland were in reality very large and the growing affluence of the rural families had to be tapped. The Hindi film industry in Mumbai was of course aware of this fact right from the beginning and has turned out the biggest hits in the history of filmmaking in the world by attractive films, which dealt with the aspirations of the people of small towns and villages. The films may not be masterpieces in the art of filmmaking but they established a mental bond with their audiences and raked in the money. India has a distinguished history of oral tradition in religion, folklore, dance, drama and music. Right while planning the mass contact campaign in our catchments area where level crossing accidents were showing no signs of improvement we realised that our conventional approach of putting advertisements in the various newspapers including in regional languages was fast turning to be ineffective and expensive. With the limited resources available to the safety branch, usually considered a pesky nuisance by the other branches at times of peace we had to do something very quickly and cheaply. One of my colleagues in Kharagpur division, who had an earlier exposure to public relations, hit upon the idea of street corner plays in the backdrop of a particularly ghastly level crossing accident in a very lightly worked section near the sea resort of Digha. On a particular festival day morning, a family of ten including the owner driver of one of those contraptions earlier described was hit by the morning tourist train from Kolkata. All but the youngest son of eleven years, perished. We took this real life example, still an emotive issue to build a script. Since then in the last three months we have taken this street play to at least twenty places, and the response of the locals has been positive beyond our wildest dreams. Touchwood there has been no report of any incident in the areas covered.

2. A mass contact campaign requires covering a very large area and repeating the message endlessly in

different forms for the contents to be absorbed and not rejected on grounds of monotony. In course of our interaction with our audiences we had asked them about our advertisements in newspapers, and realised that the recall factor was zero. Even the educated persons could vaguely recall seeing something sometimes back. We choose our newspapers very carefully, avoiding urban biases, focussing on circulation of at least a hundred thousand, and ensuring that all regional languages in our operational area get covered. The funds have never been a constraint in these areas thanks to the enlightened policies of the IR. We clearly had to look for something else. It is then that we hit upon using the cellular phone as a means of sending SMS messages. All of you must be aware that a communication revolution is sweeping through India. It is way ahead of all developed countries in terms of cellular phone penetration and next only to China. Our enquiries revealed that the BSNL had the best rural network if not in terms of quality at least in terms of reach. We also realised that majority of their 30 million subscribers spread over the three states of W.Bengal, Orissa and Jharkhand were rural people, truckers, brick kiln workers, agricultural and construction workers etc, in short our ideal



PROCEEDINGS 116/208

audience. The highly competitive cell phone environment did the rest. We negotiated a price of less than half a cent per message, and shot off our 30 million messages in little over two days. The reactions were instantaneous. The message had to conform to 160 characters and therefore was kept very curt, some of our colleagues in other zones called in to say that it was rude and shocking. It started with the reminder that the individual was responsible for his own life while traversing level crossings, a basic fact that even railwaymen had forgotten. I had gone to the remotest corner of India almost at the China border in Bomdila, and even there, people were talking about the SMS. We were elated that we had started something big. This SMS campaign was repeated every ten days till even my very old and dotty aunts could recollect about level crossing safety. 9 million messages, the feedback was encouraging and level crossing incidents did come down.

3. A mass education campaign should aim at all sections of society specially the youngsters. The SMS

campaign and the street plays were aimed respectively at the working population and the housewives and elderly men in the villages. The turnout in the street corner plays convinced us that in a young nation like in India no public awareness campaign could afford to ignore the school going children. The pre school toddlers did see the street plays and clapped as loudly as everybody else; probably they would carry fond memories of the street play into their adulthood, but middle school children were all away to school and here a gap was being created. This led to the next idea that the concerns of safety of the layman near to the railway tracks had to be addressed careful study of the school books made us realize that almost all states had some teaching material on road safety but none on railway safety. This was very surprising in a country where a number of people equal to the whole population of Australia use the services of the railways for their personal transportation needs everyday. Probably this had something to do with the height of economy being controlled by the public sector and the dour mindset of the railway man. However we got into the act and circulated course material to all the education boards namely that of W. Bengal, Orissa and Jharkhand. The sates of Orissa and Jharkhand are well on the way to incorporate this matter in their textbooks probably from next year. The state of W. Bengal however seems to be requiring a little more persuasion. Incorporating these in the textbooks would ensure that at least the children act as a catalyst for modification of the risk taking behaviour of the adults. This ‘targeting the child technique’ is used by the FMCG companies to persuade their parents into expense. We thought that we could take a leaf out of the books of the marketing men.

4. An effective mass education campaign must quickly rake the benefit of the modes of entertainment in

vogue. Changing technology ensures that the methods of entertainment also change quickly. In India for instance DVD’s were a hot favourite even before some of the developed countries had graduated fully from VHS format. Satellite T.V is the in thing now in towns but digital and analog cable transmission is the pattern in the villages and small. These individual cable operators have a movie channel where they play latest blockbusters all through the day. The cost of taking a 20 seconds spot in a commercial channel is very high and here we are competing with organisations with very hefty budgets in search of markets and tax relief. We have no such necessity or method available to us. We went once again to the small is beautiful way, and approached twenty of these small cable operators, appealing to their sense of community interest, and succeeded. We had failed to convince the big players who would not budge an inch to address social interest issues without charging big money at commercial rates. It came as a great relief that the small entrepreneur responded. We are now transmitting twenty seconds jingles into the living rooms every afternoon and evening in the nearby towns, and have literally seen awareness grow. We are in negotiation with popular FM channels. This is of slightly limited reach being constrained within 30 kilometres of major towns and cities. But the FM Channels are popular with youngsters on motorcycles and truckers who spend the night in the truck terminals. This is a target audience and therefore is of much use to us. As of now the hefty rates asked for by the FM channels are under negotiation.

5. A mass education campaign must aim at all such situations where people converge in large numbers

or where people meet and discuss their own issues. We have made it a point to distribute hand bills, display posters, enact plays and even play pre recorded messages on a PA system, with a safety counsellor at hand to answer questions, and hold on the spot quizzes, and small inexpensive prizes. We go to village Panchayats and talk to the villagers, in the presence of village elders and the elected heads of the village council. This ensures internal negative feedback and constraints to somebody who



PROCEEDINGS 117/208

flouts level crossing safety rules about which no body outside the village is even aware of, unless it has already become an accident. We call this initiative the ‘Village Safety Initiative’ and this is doing well.

6. One of the methods of Mass awareness campaigns must be to catch people staying nearby so that

they can modify their own risk taking behaviour and change those of the other road users for the better. Our safety Counsellors have been forced to get out of the bureaucratic mode and literally go and campaign for level crossing safety at level crossings, catch errant road users and correct their wrong perceptions and uninformed risk taking behaviour. One of the early realisations during our level crossing studies was that the perception of speed of a bystander being related to the rate of change of the angle subtended at the eye was usually much lower than the actual speed of the train, until the train was almost upon him. We have repeatedly made it a point in our mass campaign to emphasise that the train travels much faster than what one thinks. Even twenty years ago because of the cultural values imbibed during the freedom movement and a long history of Gandhian abstinence theories, drinking was considered unacceptable among the majority of the Indian people. Only the very rich Europeanised Indians and the very poor were seen to be drinking alcohol. But with contacts with the outside world increasing, more disposable incomes in the hands of a nation of young workers and virtually nothing being taboo in the Hindu religion alcohol use has spread across the whole population and is slowly tending to abuse. This is a problem that is repeatedly causing road users to take such risks at level crossings, which they would not otherwise take. This is a larger socio legal issue, which we have to grapple with. Similarly in the bigger cities fast air conditioned cars with tinted glasses and loud rock playing in the car stereo on a rainy night have seen quite a few accidents. We have therefore added, “listen” to our stop look and go routine. Probably rolling back the window and turning the stereo down have to be added.

7. Visual impact is one of the most important aspects of a mass education campaign. We do put the drab

regulation notices at the prescribed places near the level crossings. However we have put up on our system semi permanent plastic bright posters that are slightly gory in detail and designed to catch attention before the point of no return. We are also displaying the mobile and fixed telephone numbers of the nearest stations at the level crossings so that possibly an accident can be saved if a vehicle gets stuck in the level crossing.

The idea of this paper is not to understate the usefulness of infrastructure or redesign and maintenance of level crossings about which other speakers would be speaking in some detail in the seminar. The idea behind this paper is to suggest such means and methods of public education which in the short term have been found to be useful by us on the South Eastern Zone of the IR, and which would arrest if not reverse the trend of increase in level crossing accidents which appear to be a problem in almost every major railway system. We railwaymen are an inward looking set of dedicated professionals who always tend to look inside them selves for solution to all problems technical and managerial. However in level crossings we are subject to the vagaries of a set people whose thinking and movements have to be closely watched and changed by us, even as they are not under our control. The inward looking habits of the railwaymen have further convinced the lay public especially in India that a level crossing accident is a railway accident and all responsibility and expenses for sorting out the level crossing issues lies with the Railways. This does not take into account that the road user is surely infringing what is the essential right of way for the train. The deviant behaviour of the road user has caused serious safety hazard for the railways. The train has been involved in an accident, because the road user has not been cautious enough. In all such cases of level crossing accidents where the responsibility lies with the road user should be classified as` road accidents causing damage to railway property and endangering safety of the railways. The mass education campaign is also aimed as a means to convince the public about this fact. The education campaign approach will surely supplement and enhance the other approaches being put up in the symposium.



PROCEEDINGS 118/208


Innovative and Cooperative Leadership to Improve Sa fety at Australian Level Crossings Author(s): Phil Sochon Job Title: Deputy CEO Company: Australasian Railway Association Country: Australia

Resume of Speaker

Phil has an extensive background in different modes of transport safety including operations management, policy development and implementation and strategic planning through various roles in the Royal Australian Navy and the Maritime Services Board – Waterways Authority. With expertise in occupational road safety (fleet safety) Phil has worked with the New South Wales Government in this field and his work was acknowledged as one of the most prominent safety projects aimed at improving driver and vehicle safety. Having commenced in his current position in 2004, Phil is managing industry-wide initiatives in operational and strategic rail safety issues. He is actively involved in development of the national disability standards policy and national model rail safety legislation and the subsequent rollout of legislation all jurisdictions in Australia. Phil is also a Program Chair for Operations and Safety Research in the Australian Rail Industry’s Co-Operative Research Council. In relation to level crossings activities, Phil leads the rail industry involvement in the national joint Government and Industry behavioural Strategy which has successfully functioned for two years to date. Furthermore, he leads the Industry’s level crossing strategy implementation.

Abstract

Collisions between motor vehicles and trains at rail level crossings in Australia are amongst the most severe types of road crashes. Fatalities at level crossings constitute a small proportion of deaths on roads each year. However, these collisions contribute to almost 50 percent of the rail death toll and cost Government and the Rail Industry more than A$100 million a year. The paper will narrate the approach led by the Rail Industry in collaboration with Government to improve safety at level crossings. The paper will also outline information on all projects currently being delivered under the National Railway Level Crossing Behavioural Strategy, and will specifically focus on two major projects of the Strategy: firstly, the National Road Users Survey and how its key results will be translated into actions; and secondly, the Targeted Education and Enforcement Pilot Project in Victoria and the Northern Territory. Furthermore, the paper will detail significant lessons learnt from the development and delivery of the above projects. The Australian Level Crossing Setting Road user crashes at level crossings are estimated to cost the Australian Government and Rail Industry in excess of AU$100 million per year. According to the Australian Transport Safety Bureau (ATSB), on average 37 people die at railway level crossings in Australia annually. There are approximately 100 level crossings collisions each year between vehicles and trains. Of these, about 60 percent are pedestrians.



PROCEEDINGS 119/208

In comparison to the Australian road toll, this is minor BUT with the potential for a collision to be catastrophic in terms of loss of life and economic impact, level crossing collisions are the Australian rail industry’s number one priority. This is exemplified by two recent collisions – one in 2006 at Lismore in Victoria between a freight train and a truck which resulted in over $20m damages and a collision in 2007 between a truck and passenger train at Kerang also in Victoria that resulted in the loss of 11 lives and 20 significant injuries. In 2002 for the Seventh International Symposium on Railroad - Highway Grade Crossing Research and Safety, a national stocktake was conducted to identify the number of level crossings in Australia. With collaborations from rail regulators and accreditation authorities, the approximate figure of 9400 has since been employed to numerically represent Australian level crossings. However, this number does not include pedestrian, private, occupational, sugar cane or grade separated crossings. It is therefore assumed a much greater number of level crossings are located on Australian railways. In an effort to produce a more accurate illustration of Australian level crossings and the size of the level crossing safety issue, the Australasian Railway Association (ARA) is currently collaborating with all track managers to update this data. The national level crossing stocktake also documented the type of protection provided at each public road - rail interface. Of the 9400 public level crossings, 28% are fitted with active protection whilst the remaining sites are passively protected by Stop or Give-way signs. Following the Kerang collision the public and members of Parliament have become very vocal in calling for active protection namely lights, booms and bells at all level crossings. However, an ATSB study that examined 87 fatal level crossing collisions between 1988 and 1998 found that 51% of Australian level crossing accidents occur at actively protected sites. This decade-spanning study emphasises the fact that infrastructure alone is not the answer to increased road user safety at level crossings. Instead, the ARA believes a combination of education (aimed at changing road user behaviour), enforcement and engineering measures will lead to safer level crossings in Australia. Australia is run by a federal Government system. Under this approach, the management of level crossing safety is divided between individual jurisdictions. The country consists of six states and two territories. Each jurisdiction has a Level Crossing Safety Steering Committee that manages level crossing safety initiatives in their respective areas. Today, cooperation exists between road and rail stakeholders. However, historically, rail and road authorities from each jurisdiction worked independently to increase safety at level crossings. The following factors are attributed to this divided past: • Level crossing fatalities are a only a small portion of the national road toll but contribute to the majority of

the rail toll • The size of the Australian level crossing safety issue is artificially split in two as rail fatalities at level

crossings are included in the rail toll while motor vehicle and pedestrian fatalities at level crossings are totalled in the road toll. This method of reporting unfortunately diminishes the level crossing safety issue in Australia making it more difficult for the rail industry to highlight and communicate the size of the issue to governing bodies.

• Ambiguity existed across the Australian road and rail industry in terms of who was responsible for level

crossings. This was due to the fact that the only section of the road system road authorities were not responsible for was level crossings. Road and infrastructure maintenance at level crossings was the rail stakeholder’s responsibility however in some jurisdictions, this responsibility was unclear. To address this, the Australian Government recently introduced Interface Coordination Agreements (ICA’s). The new legislation will be explored further in this paper however, it should be noted that ICA’s now ensure road and rail authorities collaborate in their level crossing responsibilities and safety measures.

Changing Trends The ATSB recently conducted an investigation into 12 level crossing crashes between motor vehicles and trains that occurred in Australia between April 2006 and December 2007. The Government review and accompanying Safety Bulletin was initiated in an effort to publicly highlight Australia’s level crossing issue. A total of 19 lives were lost with 60 people injured in the 12 collisions examined. In almost every incident, it was found the vehicle driver failed to comply with traffic control measures. Fatigue, complacency, sighting problems, distraction and the expectation not to encounter a train were identified as underlying factors contributing to



PROCEEDINGS 120/208

motorist’s failure to stop and give-way to trains. Noting a train driver’s inability to minimise or avoid collisions due to train operational limitations, the study highlighted the need for other level crossing safety measures, notably behavioural programs rather than solely relying on level crossing safety infrastructure. The level crossing Safety Bulletin was produced in conjunction with the Australian Trucking Association and the ARA. It identified an increasing trend of heavy vehicle (truck) level crossing accidents and a consequently heightened fatality risk and death toll. Nine of the 12 accidents investigated involved heavy vehicles. This trend represents Australia’s growing freight industry. Trucks hauling three trailers were approved for rural regions of Australia in 2007 as part of the B-Triple Network. The ARA has been very vocal in its dispute of the Government endorsing B-Triple trucks however to date this has fallen on deaf ears. These trucks measure up to 53.5 meters in length and can take up to 71 seconds to clear a level crossing from a complete stop. Most level crossings located on the B-Triple network have not been appropriately assessed for such vehicles and thus, the Government’s decision is not only increasing risk at rural level crossing but contributing to the rising trend of heavy vehicle collisions at these sites. Work is underway to re-assess the sight distance required in these circumstances. The National Railway Level Crossing Behavioural Coo rdination Group (BCG) In response to major level crossing collisions across Australia and growing recognition of the difficulty of the issue, in 2003, the Australian Transport Council of Ministers (ATC) approved the National Railway Level Crossing Safety Strategy. This Strategy aims to address the complex road user, pedestrian and train interface safety issues at Australian level crossings. The strategy’s objective is to reduce the number, cost and trauma of crashes between trains and any road user in the most cost-effective means. To implement this strategy and its accompanying action plan on a national scale, the ATC formed the Australian Railway Crossing Strategy Implementation Group (ARCSIG). Composed of road and rail authorities, Government members and the ARA, the ARCSIG’s brief encompasses railway level crossing safety, rail infrastructure and level crossing policy. As a member of the ARCSIG, the ARA was tasked with implementing behavioural initiatives to increase safety at level crossings. To improve the execution of these behavioural programs, the ARA in conjunction with senior Government representatives proposed the creation of the National Railway Level Crossing Behavioural Coordination Group (BCG). Endorsed by senior Government road and rail authorities, the ATC approved the proposal for the national group in June 2006 for a two year period. Made up of senior road safety and rail safety representatives from each state and territory, the BCG represents a first within Australian transport as Government road and rail senior managers from all jurisdictions work together with the rail industry towards a coordinated national approach to improving road user behaviour at level crossings. The BCG oversees the development, delivery and evaluation of research and projects that aim to improve road user behaviour by implementing education, awareness, enforcement and technological initiatives. The group carries out its projects within a budget of $400K per annum. Of this budget, $120K was contributed in-kind by the rail industry whilst each jurisdiction contributed amounts proportionate to their population. The BCG partnership initiated a meaningful dialogue between industry and the Australian Government and has seen a high level of cooperation and collaboration between road and rail on the level crossing safety issue. This publicly highlighted the topic and built confidence within the Australian Government and public that the rail industry is serious about increasing safety and decreasing level crossing accidents. Before the BCG, jurisdictions and industries worked individually but now, resources, funding and initiatives from across Australia are pooled as the group effectively and efficiently works together towards level crossing safety. The collaborative development of behavioural programs between jurisdictions produces nationally conducted projects at a fraction of the cost of undertaking these projects in each jurisdiction. Furthermore, some jurisdictions are able to benefit when they would otherwise have not had sufficient funds. The BCG has been highly successful in the Australian level crossing environment. It has implemented a number of behavioural initiatives that are the first of its kind within the country. Key deliverables of the group are a national level crossing survey, an education and enforcement pilot, a website, an inventory, a communications package and a national workshop. These are detailed in the following.



PROCEEDINGS 121/208

1. National Rail Level Crossing Road User Behaviour al Study In 2007, the BCG commissioned the first National Rail Level Crossing Road User Behavioural Study across all jurisdictions. Three focus groups and 25 in-depth interviews were followed by a quantitative survey of over 4400 road users across Australia. The study identified self-reported behaviours and attitudes to measure participant’s awareness, knowledge and perceptions of the rules and risks associated with level crossings. It also measured how respondents view and interact with railway level crossings. Participants were road users aged 18 years and older in possession of a current driving licence who had travelled over a level crossing at least once within the past six months (exclusive of being a passenger). The study intentionally over-represented regional and rural Australians in an effort to mirror the prevalence and location of Australian level crossings that are predominantly found in rural and regional Australia. Some of the key survey takeouts follow: • 24% reported engaging in illegal usage of a level crossing one or more times. This included crossing when

a train was visibly approaching, not stopping at a stop sign, accelerating to pass under a lowering boom gate, not waiting for the lights and boom gates to cease operation before proceeding across train tracks, avoiding the boom gate by driving around it and becoming trapped between lowered boom gates in their effort to rush across a level crossing.

• Approximately one in five acknowledged they had travelled over a level crossing and not realised until after

they had crossed • One in five were not aware of any penalties for breaking the rules at level crossings while 66% believed

they were less likely to be penalised for breaking rules at level crossings than driving at speeds exceeding the speed limit

• Driver inattentiveness and impatience were collectively identified as the greatest factors contributing to

increased risk at level crossings • One in four reported engaging in risky behaviour at level crossings yet not all participants classified

crossing when a train is approaching as risky • 16 – 25 year old drivers were identified as the group most at risk at level crossings. Interestingly, this group

was self-aware of their heightened risk yet older drivers were less aware of their own risk. The findings from this innovative national survey are available to each jurisdiction and are now being utilised in conjunction with the education and enforcement pilot (detailed below) to tailor a national communications package. 2. Exemplar Education and Enforcement Pilot Also in 2007, the BCG conducted the Exemplar Education and Enforcement Pilot. A ‘before and after’ study, the program measured road user behaviour at level crossing sites in metropolitan and rural Victoria and examined the results of a mining company’s level crossing initiatives in the Northern Territory. Behaviour at the nominated level crossings sites was monitored to measure compliance with stop and give-way signs as well as crossings fitted with lights only and those fitted with lights and booms. A local education campaign and accompanied enforcement was then conducted in these areas for four weeks. Following this month long period, behaviour was remeasured at the test sites and control sites situated in others parts of Victoria to determine the effectiveness of the education and enforcement programs. The pilot program was conducted to provide a basis on which to develop guidelines for effective, practical and sustainable enforcement programs as well as an associated community education program for implementation at level crossings across all jurisdictions. The aim of the pilot was to provide a basis on which to develop guidance materials for rail organisations to engage enforcement agencies in level crossing safety. It also



PROCEEDINGS 122/208

aspired to create resources to guide rail organisations to engage with their local road safety agencies to conduct community awareness campaigns. This study recognises the fact that increased compliance with road rules at level crossings will decrease the number of level crossing collisions and that enforcement is a critical element of heightened level crossing safety. The ARA participated in the program with the Public Transport Safety Victoria (project manager) and VicRoads. Three Community Road Safety Councils delivered local education campaigns whilst Victoria Police conducted enforcement at the sites. The pilot demonstrated that enforcement has a positive effect on road user compliance with Stop signs at level crossings. A mining company’s compliance program in the Northern Territory was examined as a case study. After unsuccessfully stationing a security guard at a level crossing to enforce stopping at the Stop sign, the company introduced a log book that drivers had to exit their trucks to sign before they travelled over the crossing. The book remained for four months before being removed and replaced intermittently with security. Over three months of monitoring, non-compliance was almost eliminated and remained very low over the following five months with only the occasional heavy vehicle failing to comply (the most recent information available). As the sample size of the Victorian pilot was relatively small, it is believed this project will be used as preliminary work. However, it does suggest very low levels of compliance at crossings and the potential benefits of controlled education and enforcement to improve safety. Findings from the project will be incorporated into the communication package to ensure the resource is an effective tool for educating the public and communicating the level crossing safety message. 3. Inventory The BCG is compiling an inventory of Australian and overseas level crossing behavioural programs. Details of targeted safety education programs, mass media campaigns, enforcement programs and community awareness campaigns are being collected and put into a specific location on the ARA website. The inventory will be updated regularly and will act as an information and reference source for those interested in level crossing initiatives that have been conducted in Australia or overseas. The inventory will also provide points of contacts for further information regarding specific campaigns or programs and Australian level crossing statistics. 4. Website Level crossing safety web pages containing relevant information on level crossings and safety behavioural initiatives are located on the ARA website. These pages are continually updated with key deliverables of the BCG and other ARA level crossing projects. Links to useful Australian and overseas level crossing websites is also provided. 5. Communications Package A Communications Package is being created by the BCG using the findings from the National Rail Level Crossing Study, Exemplar Education and Enforcement Pilot, and awareness and education programs from overseas, as baseline information. The communication package will consist of a resource kit that will include print materials and radio and television scripts that will be available to all jurisdictions to provide nationally consistent communication materials for future level crossing safety communication campaigns undertaken by Government and industry. Whilst users will be required to fund the distribution of the communication materials, the availability of such resources will save time and money whilst providing effective communication tools that are easily and quickly obtainable. The creation of this package will help distribute consistent and effective safety messages to the appropriate target groups as identified within the national survey and education and enforcement pilot.



PROCEEDINGS 123/208

6. National Workshop On 21 February 2008, the BCG in collaboration with the ARA held the “Safer Road User Behaviour at Level Crossings – A National Workshop”. The National Workshop presented key initiatives completed to date by the BCG and provided an environment for stakeholders to discuss future level crossing safety options. During the Workshop, attendees formed groups and participated in an interactive session brainstorming ideas and future BCG projects to reduce level crossing collisions and injuries as well as safety initiatives that could be implemented by groups outside the BCG. Senior representatives from Commonwealth and State Government authorities, the rail and trucking industry, rail regulators and investigators, unions, enforcement agencies and the Victoria State Coroner’s Office participated in the workshop which was labelled a success by the ARA, BCG and attendees. 7. The future of the BCG It is proposed the BCG will continue with a focus on these activities: • Conduct a detailed national study regarding vulnerable road users (specifically pedestrians, cyclists and

people with disabilities), and their use of level crossings (or adjacent to level crossings • Develop and evaluate a communications campaign aimed at vulnerable road users at level crossings • Develop and evaluate nationally available level crossing education materials for upper primary and junior

secondary schools • Stage a national demonstration project linking education, engineering and enforcement aimed at the heavy

vehicle industry and their use of level crossings • Use the Objective Data, List of Solutions, and Addressed Plans of Action (OLA) approach that employs

multi-lateral collaboration to work with companies to increase safety at level crossings • Evaluate projects initiated during the first two years of the BCG (BCG1) • Work with National Transport Commission to develop, implement and evaluate a national protocol

regarding enforcement at level crossings • Identify findings from BCG1 projects that are conducive to the application of ITS or other engineering

initiatives and prepare a report to guide level crossing engineering standards and practices improvements • Provide input to research programs and projects related to road user behaviour at level crossings • Evaluate the Operation Lifesaver program with the view to adopting culturally and behaviourally appropriate

tools in the Australasian context • Continue enhanced communications including a national workshop and regular update of the website and

inventory New Initiatives In addition to the national work of the BCG, a number of other initiatives are underway.

1. Victorian actions following Kerang disaster The state of Victoria has the highest density of level crossings in Australia and has sadly seen a significant number of high profile level crossing events in the last year. The Victorian State Government has initiated a number of key improvements in level crossing safety and in 2007 announced a $33.2 million package to improve level crossing safety in Victoria. This package includes the following: automated advance warning signs, rumble strips, increased enforcement, an advertising campaign, line of sight improvements and the trial of red light cameras at level crossings.



PROCEEDINGS 124/208

• Automated advance warning signs : installed at 26 level crossings on highways and 27 sites with

high traffic volumes. Flashing signs will be installed approximately 250 meters before a level crossing. Activated by approaching trains, these will visually warn road users of the impending level crossing and draw attention to road signs.

• Rumble strips: installed at 200 level crossings across the State. These will provide a 250 metre audio,

sensory and visual warning to drivers approaching level crossings. Tragically, on Easter Monday this year a family car collided with a train on a regional level crossing that had recently been upgraded to include rumble strips. The mother and daughter died in the accident while the father and two sons survived. This accident raised a public query about the effectiveness of rumble strips.

• Red light camera trial: at two busy levels crossing locations in Victoria, cameras will be trialled for 12

months. On completion, the effectiveness of the cameras will be determined and the possible implementation of red light cameras at other level crossing locations decided upon.

• Road speed reductions: are being introduced at 72 level crossings in Victoria. The speeds have been

lowered from 100km/hr to 80km/hr providing drivers with the ability to not only better assess situations but react and stop more quickly if necessary.

• Increased infringement penalties: Australian level crossing infringement penalties differ from State to

State. The Victorian Government set the benchmark by boosting the $177 fine and three demerit points to a $430 fine and the loss of four driving licence points (drivers of a full Victorian licence have 12 points in total). They also introduced fines for speeding to beat a train, crossing train tracks when the lights and bells are activated or driving around lowered boom gates. Other States and Territories have raised penalties.

2. Interface Coordination Agreements (ICA’s)

Prior to ICA’s, road and rail authorities worked independently in their efforts to combat risk at level crossings. ICA’s call for rail infrastructure track managers and relevant road authorities (state, commonwealth and local council authorities) to join forces to identify potential risks at individual level crossings. The agreements require the combined creation of one or more plans to combat identified risk at each level crossing. Not only do ICA’s provide an environment to further manage risk at level crossings, they also ensure the road and rail industry work together to formulate measures that manage and alleviate identified risks at each site. Local councils are required to monitor ICA’s to ensure both parties maintain their parts of the agreement. The legislation calls for periodic formal reviews guaranteeing the risk management plans remain practical and consider any changes to or near-by level crossing sites.

3. Rail Industry Level Crossing Strategy In 2007, the ARA developed a National Strategic Plan indentifying seven objectives for Industry. Objective five aims “to bring about the introduction of measures that substantially reduces level crossing collisions”. To achieve the anticipated outcome of reduced level crossing fatalities, injuries and associated costs, the Industry Level Crossing Strategy was produced. The strategy recognises that level crossing crashes are a low priority issue for road authorities and acknowledges the need for the rail industry to lead the way in level crossing safety initiatives. The main focus of the 2008/09 action plan will be education, enforcement and engineering. This will incorporate a wide range of measures whereby industry will work with Government and other stakeholders to reduce level crossing fatalities and injuries.

4. Intelligent Transport Systems (ITS)

In February 2008, in collaboration with ITS Australia, the ARA convened the first national ITS Workshop which aimed to promote the uptake of Intelligent Transport Systems (ITS) as a means to increase safety at level crossings, achieve more effective road user behavioural controls and reduce cost of infrastructure. Participants included experts from road and rail industries and Government as well as researchers and technology



PROCEEDINGS 125/208

companies. The inaugural ITS level crossing workshop was labelled a success by attendees, ITS Australia and the ARA. The Workshop provided an opportunity for participants to examine a range of technological issues and initiatives as it aimed to stimulate interest in ITS by presenting the options available to improve road and rail safety at level crossings. Numerous technologies were showcased during the day including: • A red light monitoring system that is able to predict whether a vehicle will stop at the flashing lights of a

crossing. • A signal intercepting unit vehicle fitment that transmits a clear spoken warning alert to drivers at a

predetermined distance of upcoming level crossings. A Working Group has been formed of ITS Australia, the Victorian Transport Association, senior representatives from Government and the ARA to progress ITS uptake by the road and rail industry as a means to increase level crossing safety.

5. Intelligent Access Program (IAP)

The ARA is currently exploring the use of GPS tracking technology as a measure to oversee the operation of heavy vehicles at level crossings with Transport Certification Australia (TCA). TCA have developed the GPS-based device that interfaces with a monitoring mechanism to establish if any unsafe behaviour has occurred. Introducing IAP into trucks, particularly the B-Triple road trains will increase safety at level crossings as the behaviour of trucks at level crossings can be monitored. This will assist in ensuring that trucks, most notably B-Triples operate safely at level crossings.

6. Research

The Rail Industry has established a Cooperative Research Centre with various leading Universities in Australia. Within the research program are a number of important level crossing projects, namely the Survey of Railway Level Crossing Safety Research a major project to utilise motor vehicle simulators to establish relevant human factor components leading to level crossing collisions. Another project will explore new low cost level crossing protection systems for crossings in regional areas and occupational crossings with high speed passenger trains. Conclusion The BCG has been extremely successful at implementing an array of research projects and initiatives to increase level crossing safety. It has demonstrated how jurisdictional Governments have been able to work very effectively with Industry to produce cost-effective behavioural outcomes. Clearly there is much to do in this area and it is hoped the next instalment of the BCG is granted Government approval. The BCG has brought jurisdictions together, pooling their funding and resources so Australia works towards safer level crossings as a nation rather than separate entities. The Australian rail industry believes every level crossing fatality can be avoided and is working in collaboration with the BCG and other level crossing initiatives to achieve zero deaths at level crossings.



PROCEEDINGS 126/208

Session 7 - Engineering and Operation Chairman: Philippe Feltz, RFF, France

Vision Zero – Guiding safer level crossings

Author(s): Paulina Holmberg, Helena Hook, Lena Eklund,

Olle Mornell

Job Title: Road engineer, team leader

Company: Swedish Rail Administration and Swedish Road Administration

Country: Sweden

Resume of speaker Paulina Holmberg has a Master of science in civil engineering and has worked at the Swedish road administration consultant services since 2002. Traffic safety is Paulina’s main task and she coordinates the Swedish Road administration’s OLA activities (acronym for: Objective data, List of solutions, Addressed action plans). The interest of level crossing safety was raised when Paulina chaired in Level Crossing OLA which started in 2005. Abstract Vision Zero is conceived from the ethical base that it can never be acceptable that people are killed or seriously injured when moving within the transport system. System designers are responsible for the design, operation and the use of the transport system and are thereby responsible for the level of safety within the entire system. On the basis of Vision Zero the Swedish road administration developed the OLA working approach in 2002. Two severe accidents in 2004 and 2005 brought focus on the dangers of level crossings. Banverket took the first steps to cooperation among key players. Vägverket provided the OLA working approach. The decision that Banverket and Vägverket should pursue Level Crossing OLA together was made in the spring of 2005. Level Crossing OLA resulted in action plans from all participating key players. The results also provides unique collaboration and measures that one key player alone wouldn’t be able implement. Without Level Crossing OLA these key players wouldn’t have assembled and collaborated during these systematic circumstances. Cooperation between key players is essential when working towards Vision Zero. Working models, such as OLA, ensure a systematic approach. Introduction Vision Zero is conceived from the ethical base that it can never be acceptable that people are killed or seriously injured when moving within the transport system. Cooperation between key players is essential when working towards Vision Zero. Every key player’s contribution to improve safety in the transport system is needed. A tool to make the cooperation systematic does the work easier.



PROCEEDINGS 127/208

Background For many years the emphasis in traffic safety work has been in trying to encourage the road users to respond in an appropriate way, typically through licensing, testing, education, training and publicity to the many demands of a man-made and, increasingly, complex traffic system. Traditionally the responsibility for safety has been placed on the user rather than on the designers of the system. The “Vision Zero” was passed in the Swedish Parliament 1997 and was the start of an entirely new way of thinking regards traffic safety. The system designer was given the primary responsibility for the safety in the transport system. Vision Zero has also been assumed in other countries throughout the world. Vision Zero is conceived from the ethical base that it can never be acceptable that people are killed or seriously injured when moving within the transport system. The long term goal is that no-one will be killed or seriously injured within the Swedish transport system. Vision Zero accepts that preventing all accidents is unrealistic. The aim is to manage them so they do not cause serious health impairments. The long term objective is to achieve a transport system which allows human error but without it leading to serious injury. The keystone for Vision Zero is to emphasise on how the entire system can operate safely rather than the ability of the individual user. System designer has primary responsibility

• System designers (e.g. road and rail administration, vehicle manufacturer, police, interest organisations) are responsible for the design, operation and the use of the transport system and are thereby responsible for the level of safety within the entire system.

• Users are responsible for following the rules for using the transport system set by the system designers. • If the users fail to comply with these rules due to lack of knowledge, acceptance or ability, the system

designers are required to take the necessary further steps to counteract people being killed or injured. Strategic principles The traffic system has to adapt to take better account of the needs, mistakes and vulnerabilities of users. The level of violence that the human body can tolerate without being killed or seriously injured forms the basic parameter in the design of the transport system. Vehicle speed is the most important regulating factor for safe road traffic. It should be determined by the technical standard of both roads and vehicle so the level of violence that the human body can tolerate doesn’t exceed. In rail traffic it is necessary to provide a safe system and reducing risk because accidents often have catastrophic implications. The most important factor is technical solutions eliminating human errors carried out by personnel in the rail traffic system. Level crossings are interface between two infrastructures, rail and road. Accidents that occur in level crossing are mainly fatal and the strategy is to prevent them. Visualizing Vision Zero



PROCEEDINGS 128/208

Figure 1. The first picture illustrates how we normally consider a level crossing. The crossing has half-barrier protection and good visibility.

Figure 2.

The second picture illustrates how a level crossing can affect road users and cause injuries or fatalities. The first thought might be that the road user is speeding and therefore responsible for the implications. But blaming the victim is not acceptable with the vision zero approach. The question is instead: What can the system designer do to prevent this unwanted situation? There are several solutions as installation of technical protection systems, visibility improvements or education.



PROCEEDINGS 129/208

Figure 3.

The third picture illustrates how a level crossing can affect rail users cause injuries or fatalities. The first thought might be that the road user shouldn’t have chosen this road because of the type of vehicle. However the same question as before is justified: What can the system designer do to prevent this unwanted situation? Witch leads to other questions: Did lorry driver’s employer give the driver a reasonable chance to plan the route? Has the rail infrastructure manager caused an obstacle for the trailer by changing the track position without contacting the road infrastructure manager? This discussion makes it clear that system designers need to cooperate in order to improve safety in level crossings. How to do it systematically? One way is to use the OLA working method where key players meet and, from a basis of facts, focus on solutions to solve the common problems and finally announce addressed action plans. Material and method Level Crossing OLA On the basis of Vision Zero the Swedish road administration developed the OLA working approach in 2002. OLA is an acronym for Objective data, List of solutions and Addressed action plans. The working approach involves system designers to work together to provide solutions to a common problem. With this approach, all key players are offered an opportunity to present desired measures they are able to implement and as a result to contribute to improved traffic safety. Two severe accidents in 2004 and 2005 brought focus on the dangers of level crossings. Banverket took the first steps to cooperation among key players. Vägverket provided the OLA working approach. The decision that Banverket and Vägverket should pursue Level Crossing OLA together was made in the spring of 2005. Eight organizations took part in the collaboration:

• Association of Swedish Train Operators • Bombardier Transportation • National Federation of Private Road Associations • Stockholm Transport • Swedish Association of Local Authorities and Regions • Swedish Association of Road Haulage Companies • Swedish Rail Administration • Swedish Road Administration



PROCEEDINGS 130/208

O - Objective data In the first phase facts are presented and discussed. The facts can for example be:

• Accident statistics • In-depth studies of fatal road accidents • Facts and knowledge from other key players

By studying the chain of events point by point, system designers can together offer an idea of why the accident became fatal. Some issues include: Is the road and rail environment suitably designed? Was appropriate safety equipment used? What was the cause of death? In Level Crossing OLA the Objective data consisted of information about level crossings in Sweden and Level crossing accidents statistics from road and rail administration. There are approximately 10,000 level crossings in Sweden. About 3,000 of them have some kind of signalling equipment. Banverket is responsible for 8,000 level crossings. Of these, 2,200 have barriers and 700 some kind of light and/or audible signal. All installations are automatically operated by the movements of trains. The number of grade separated intersections is nearly 3,000. Every year some one hundred crossings are closed. At this rate, level crossings will remain a reality for many years.

Figure 4. Level Crossing Accidents in Sweden 1996-2007

A study of in-depth studies of fatal level crossing accidents 1997-2003 shows that seven of ten people were killed in accidents on the private road network and that more than four of ten people were killed at level crossings without road protection installations.



PROCEEDINGS 131/208

Figure 5. Road manager. Level crossing accidents 1997-2003

Figure 6. Protection Installations. Level crossing accidents 1997-2003

L - List of solutions/actions In the second phase system designers present and discuss proposals and ideas for solutions - both in short and long term. Measures can be both large and small. It is important that discussions are forward looking and focus on finding opportunities for improvement. The discussion should mainly concentrate on what each system designer can do, either individually or together with another system designer. A - Addressed action plans The third phase concerns what concrete measures system designers can initiate to improve safety. A declaration of intent is a description of what, when and to what extent measures will be implemented along with the aim of each measure. Announcement When the action plans from participating system designers are signed by the director-general or CEO the result of the OLA-project goes public. The OLA is also documented and published on a website, www.vv.se/ola. The result and addressed action plans from Level Crossing OLA were presented at the Swedish National Rail administration’s annual safety conference in March 2006.



PROCEEDINGS 132/208

Follow-up meetings System designers are responsible for implementing and following-up their own action plans. A year after the announcement the first of three follow-up meeting takes place in purpose to share information about accident trends and implementation of the action plans.

Results

Level Crossing OLA resulted in action plans from participating key players: Association of Swedish Train Operators • Stimulating member companies, those operating at ports and industrial track facilities, to equip their

locomotives with warning lights. Target: 80% of the traction vehicles in the above operations will be equipped with warning lights. Date: 31 December 2006

• Actively promoting a uniform, systematic handling of incidents and prevailing conditions at level crossings by member companies. Target: Documented common reporting routines Date: 31 December 2006

Bombardier Transportation • Bombardier is open to work together with train operators to find suitable technical solutions so that both

new and older vehicles can be equipped with an automatic headlight flash in the event of an emergency brake. Bombardier considers these initiatives to be an important continuation of their ongoing work to improve railway safety. In addition, Bombardier can develop solutions to install the European Railway Traffic Management System (ERTMS) technology in the most common Swedish vehicles. ERTMS technology provides new possibilities to send information to the drivers directly into the drivers’ cab. The system also makes it possible for the level crossing control equipment to automatically brake the train if needed.

• Bombardier can develop products for protection of level crossings using ERTMS technology. Phase one will cover regional lines, where a large number of unprotected level crossings are present. ERTMS technology reduces the amount of equipment required and simplifies the interface to the railroad signalling system. This in turn results in reduced installation and operational costs compared to the technology used today. In the future, similar systems can be developed for main lines. Bombardier also sees possibilities to build cost effective, stand alone level crossing protection systems by using new computerised signalling products in new and more effective ways. The new technology reduces the need for long cables, making it possible to implement remote control and remote diagnostics, and to reduce the operational cost. Cost for back up power systems can also be reduced. In cooperation with its customers Bombardier wants to look at different solutions for LED signals for road traffic and different techniques to detect cars or other vehicles within the railway’s safety zone. A further development of ERTMS regional technology makes it possible to create solutions for industrial areas such as harbour tracks and similar environments, so that the driver can both control and supervise the signals at a level crossing. Today, normally only a red flag or a sound signal is used. Bombardier can also reduce the mechanics and cabinets for level crossing control equipment so that the need for building permits and building time is reduced to a minimum. Bombardier sees cooperation with Vägverket and Banverket concerning these questions as an opportunity to perhaps alter the legislation so that in the future it may be possible to use the new technology, even if this is not yet possible today.

National Federation of Private Road Associations • Publishing a special feature issue of its members’ magazine that focuses on the dangers involved at level

crossings and the individual road manager’s joint responsibility for the surveillance of level crossings as regards equipment and sight requirements. The intention is that this information will have reached 8 000 families in 2006.

• Taking part in a dialogue project in collaboration with the Swedish Road Administration in 2006 to learn what road users experience as problems and subsequently find solutions for these.

Stockholm Transport (AB SL) • Together with the Swedish Road Administration and Swedish National Rail Administration, AB SL wants to

have the Swedish Road Sign Ordinance, SFS 1978:1001 amended so that it will be possible to use the red light installation illustrated in Figure 12.2 ”Signalling equipment at a movable bridge, ferry landing, airfield,



PROCEEDINGS 133/208

emergency vehicle exit, etc” at tramway level crossings. The installation used today, as shown in Figure 12.3 ”Signalling equipment to attract special attention”, displays a yellow light and is the cause of a great many accidents and incidents.

• Further, AB SL wants to take part in a project run by the Swedish Road and Transport Research Institute to actuate the question of being able to use a special barrier installation at tramways in Sweden. Funds have been allocated for this research for the years 2006 and 2007. These installations exist in several EU countries and it should therefore be possible to approve their use at tramways in Sweden as well.

Swedish Association of Local Authorities and Regions • In 2004 the Swedish Association of Local Authorities and Regions and the Swedish Road Administration

jointly published the 1st edition of a new planning tool,”Vägar och gators utformning” (VGU)” [“Road and Street Design”] to replace the recommendations in the former planning manual. VGU is intended to be a living document that will be regularly updated and supplemented. There could be a second edition issued in 2007. VGU should be supplemented to include the design of barrier crossings in complicated urban environments and where space is at a premium. The Swedish Road Administration is the principal for VGU in collaboration with the Swedish Association of Local Authorities and Regions. Swedish National Rail Administration should take part in further developing this section.

• Tramway lines on “street tracks” cross other traffic at-grade in many places. Larger crossings are regulated by traffic lights. Signals for pedestrians entail special problems. A marked pedestrian crossing is only legal on the part of the street intended for motor vehicles. At tracks in the middle of the street, for example, the pedestrian marking, which is regulated by the Road Sign Ordinance, cannot supersede the provisions in the Railway Act, which give precedence to rail traffic. This can be further complicated by the presence of buses within the track area. In the opinion of the Swedish Association of Local Authorities and Regions, a clear and pedagogically correct legal formulation of the Road Sign Ordinance and the Railway Act is a matter of urgency for both pedestrians and tramway drivers. The Swedish Road Administration is the principal for the Road Sign Ordinance. Banverket, Stockholm Transport, other municipalities with tramway lines and the Swedish Association of Local Authorities and Regions are the stakeholders concerned.

• Codes of regulations differ for tramway and railway traffic. Tramway traffic is subject to the terms and conditions of road traffic. Today there is no protection installation especially intended for tramway level crossings. There is, however, a demand from many municipalities that this type of equipment be developed. Although such installations exist in many different places in Europe, they have not been approved by Swedish public authorities. Stockholm Transport, with the support of the Swedish Association of Local Authorities and Regions, has formally requested that Banverket and the Swedish Road Administration initiate analyses of technical solutions, codes of regulations, road user behaviour, attitudes, etc with the intention to develop tramway protection installations approved for use in Sweden. Banverket, the Swedish Road Administration, Stockholm Transport and the Swedish Association of Local Authorities and Regions are the players concerned.

• The Swedish Association of Local Authorities and Regions intends to work for a revision of the regulations that apply to level crossings between pedestrians and railway traffic at stations with little traffic and where grade-separated crossings are disproportionately expensive and a roundabout way via another grade-separated crossing is disproportionately long. Banverket’s code of regulations for pedestrian passage at low-trafficked station areas needs to be reviewed and discussed from a safety and community development perspective. Banverket and the Swedish Association of Local Authorities and Regions are the players concerned.

Swedish Association of Road Haulage Companies (SÅ) • SÅ intends to find a suitable way to incorporate the question of safe passage through level crossings into

its work on road safety. • SÅ intends to spread any relevant information about level crossings produced in collaboration with other

parties. • SÅ intends to assist in the inventory compilation through providing incident reports obtained from members

concerning crossings perceived as dangerous. • SÅ intends to submit demands to the automotive and vehicle superstructure industries concerning speed

inhibitors, or the equivalent, for lorries with a crane in a raised position. • SÅ intends, in collaboration with other parties, to help develop routines for transport exemptions so as to

include relevant information about level crossings. • SÅ prioritises road safety in compliance with its declaration of intent and the objectives set by the Board of

Directors and intends to assign high priority to road safety in the coming years.



PROCEEDINGS 134/208

• SÅ supports road haulage companies in their work on preparing a road safety policy and developing a sustainable road safety system. Special road safety advisors have been trained and are now available for consultation throughout the country.

• SÅ has produced motivational material and concrete guidelines for the work on road safety. Road haulage companies that adopt a road safety policy can register this at SÅ and have it published on the SÅ home page. This work is an ongoing process.

Swedish Rail Administration (Banverket) • Banverket intends to abolish 400 level crossings in collaboration with road managers in the period 2006 -

2008 • Banverket intends to improve the protection at 40 level crossings in collaboration with road managers in the

period 2006 - 2008 • Banverket intends to build 60 grade-separated crossings in collaboration with road managers in the period

2006 - 2008 • Banverket intends to look for new solutions where grade-separated crossings could replace busy level

crossings in collaboration with road managers as well as inquire into co-financing possibilities in the period 2009 - 2015

• Banverket intends to begin changing over to better light signals (with light-emitting diodes LED) • Banverket intends to publish information about level crossings at the Banverket website • Banverket intends to introduce clearer routines for retaining the road profile in connection with rail track

works • Banverket intends to start to introduce a ”drive through” sign on the back of full barriers for vehicles caught

on the tracks between lowered barriers • Banverket intends to set up signs at level crossings, informing the general public where to call if the road

protection installation is out of order In collaboration with other players: • Banverket intends to identify problem crossings. Pending the development of a method to do so, data

already available at Banverket or other players will be used. • Banverket intends to encourage road managers to set up road signs as soon as possible to prohibit the

transport of hazardous goods at problem crossings. Other more long-range measures could be undertaken later.

• Banverket intends to investigate the possibility of new or improved protection installations at level crossings, such as a road central barrier where there are half barriers,

• Banverket intends to introduce breakage supervision in all barriers, to introduce obstacle detectors in additionally cases.

• Banverket intends to arrange the “European Level Crossings Research Forum” in Sweden in 2007. • Banverket intends to take part in the work on preparing regulations and recommendations, for example for

speed limits on roads, regulations for transport exemptions, advance warning signals, hanging reflectors on barriers, blue and white gantries, protection installations for tramway level crossings, etc.

• Banverket intends to prepare a method for identifying problem crossings with a high frequency of serious near-accidents in consultation with the rail companies.

• Banverket intends to prepare routines for better/mutual feedback of experience regarding accidents at level crossings together with the Swedish Road Administration in 2006.

• Banverket intends to take part in the work on supplementing VGU (”Road and Street Design manual) so as to include level crossings.

• Banverket intends to facilitate the handling of transport exemptions by the Swedish Road Administration by providing it with data about level crossings with limited accessibility.

Swedish Road Administration (Vägverket) • Vägverket intends to conduct dialogue projects in order to find out the problems experienced by local

residents and develop site-specific solutions for level crossings on the private road network. Every Vägverket region is to conduct at least two dialogue projects in 2007. The findings will be compiled in a report and sent to Banverket and the National Federation of Private Road Administrators.

• Vägverket intends to have updated the section of ”VGU”* concerning the design of level crossings by 31 December 2007. The focus will be directed at the design of crossings in rural areas. Banverket will be involved in this process. *”Road and Street Design”, VGU, is a manual for designing roads and streets. The



PROCEEDINGS 135/208

intended target group includes road and street planners as well as road managers involved at the planning, detailed investigation and design stages. The design recommendations are based on the intended function with regard to accessibility, safety and the environment, irrespective of who the road manager is.

• Vägverket intends to initiate collaboration with Banverket, the Swedish Association of Local Authorities and Regions and Stockholm Transport in order to develop and improve the regulations governing tramway level crossings. The collaboration will begin in 2006 and the review will be completed by no later than the end of 2007.

• In 2006 and 2007, the Vägverket intends to take part in a collaboration initiated by Banverket to identify level crossings that present a problem for heavy, long and low transport vehicles to drive through. The identification, which is to be completed in 2007, will result in better routines at Vägverket for the administration of transport exemptions.

• Vägverket intends to collaborate with Banverket concerning information and data related to road safety and level crossings. This information is to be published on Vägverket website by 31 December 2006. The aim is to specify important information regarding safety at level crossings. The target group includes both professional drivers and the general public. The material will be sent to professional drivers via the Swedish Association of Road Haulage Companies and be made available to the general public at the Banverket and Vägverket websites.

• In 2006, Vägverket intends to work together with Banverket to draw up routines for better/mutual feedback of experience regarding accidents at level crossings. Since 1997, Vägverket has been conducting in-depth studies of all fatal accidents that occur on roads and streets. An in-depth study is a structured collection of data to find out what caused the tragedy entailed by a fatal accident. The sharing of experience is intended to supplement the picture of the problem with facts as seen from both the road and rail manager perspective.

Two follow-up meeting has been carried out in Level Crossing OLA one in March 2007 and the second in March 2008. Some of the results from implemented action plans are presented here: • Vägverket, Banverket and the Swedish Association of Road Haulage Companies has carried out a

campaign where drivers are enforced to report, in their point of view, level crossing with poor standard. The reports can be filled in on a web site. The web site is open for every one not only lorry drivers. The result is about one report a week.

• Banverket and Vägverket arranged the “European Level Crossings Research Forum” in Sweden in November 2007.

• Banverket has, in collaboration with road managers, abolished 200 level crossings. • Banverket has, in collaboration with road managers, improved road profiles in 20 level crossings where low

vehicles can be grounded. An inventory of 2000 level crossings is carried out in 2008. • Banverket is testing signs in the inside of full barriers with the text: “Drive through the barrier!” in some level

crossings. • The National Federation of Private Road Associations has published a newspaper issue about risks at level

crossings which reached 8000 members. • The Association of Swedish Train Operators has implemented warning lights on 80% of the traction

vehicles operating at ports and industrial track facilities.



PROCEEDINGS 136/208

Conclusions

• Cooperation between key players is essential when working towards Vision Zero. Working models, such as OLA, ensure a systematic approach.

• The OLA working approach provides a short and clear process witch makes it easy to gather key players for collaboration.

• Participating in an OLA is on every key player free will which means that everyone invited do not always participate.

• When cooperating with key players it is important to organise and plan the work properly. The OLA working approach provides this and makes sure that the used data is objective.

• Every key player needs an internal process to gain approval from its organisation to carry out their addressed action plans.

• The OLA working approach includes the director-general or CEO signing the addressed action plan. This implies an aware decision from the management to implement measures regarding level crossings.

• At the follow-up meetings we have realized that more time is needed to implement some of the addressed action plans.

• An evaluation among participating key players in Level Crossing OLA shows that the OLA working approach is a successful method to assemble key players and achieve results.

• The results from Level Crossing OLA provides unique collaboration and measures that one key player alone wouldn’t be able implement.

• Without Level Crossing OLA these key players wouldn’t have assembled and collaborated during these systematic circumstances.

References

Swedish Road Administration, “Safe traffic - Vision Zero on the move”, http://publikationswebbutik.vv.se/shopping/ShowItem____1317.aspx, (2006) Swedish Road Administration, ”It must not happen again”, http://publikationswebbutik.vv.se/shopping/ShowItem____1127.aspx, (2007) Swedish Road Administration, ”Website about Level Crossing OLA”, http://www.vv.se/templates/page3____19602.aspx, (2007-03-22)



PROCEEDINGS 137/208

Session 7 - Engineering and Operation

A Technology Comparison of Two In-Vehicle Warning Methods at Level Crossings with Human Factor Implic ations

Author(s): Paul Bousquet, Steven Peck

Job Title: Systems Engineer, Mechanical Engineer

Company: Volpe National Transportation Systems Center

Country: United States of America

Resume of Speaker

Paul has worked as an engineer for the U.S. government for over 23 years, the past 8 at the Volpe National Transportation Systems Centre. His work at the Volpe Centre in the Railroad Systems Division includes supporting the FRA's efforts in vital railroad control systems and the promotion of positive train control. Paul serves as a program manger for the FRA’s Railroad System Safety research project at the Volpe Center. Paul has also provided support to the FRA in the area of intelligent systems for highway rail intersections and advanced train control. He has also worked on the Maglev Deployment Program in support of a cooperative agreement between the U.S. and German governments and currently serves as the Volpe technical lead for its support to the Federal Transit Administration’s Urban Maglev Program. Paul was a member of the IEEE Highway Rail Intersection Working Group that developed a standard for this interface. Prior to working at the Volpe Centre, Paul worked for the Naval Sea System's Naval Undersea Warfare Center in Newport, RI as a Senior Software Engineer for Unmanned Undersea Vehicles. Education: B.S. in Computer Engineering from University of Massachusetts-Dartmouth M.S. in System Engineering from Boston University

Abstract

Providing warning indications to vehicle operators approaching a level crossing can offer a valuable advanced notice to potential hazardous situations. Traditional methods have focused on signage and roadside active warning systems to alert vehicle operators. Providing an in-vehicle notification to an operator that they are approaching a level crossing can help return the driver’s attention to the road. With technology continuing to advance and the proliferation of advanced electronics in today’s motor vehicles, the capability of implementing in-vehicle warning systems for level crossings is increasing dramatically. With more of today’s vehicles manufactured with built-in displays and navigation systems, it is much easier to provide a system that will alert the vehicle operator to an approaching crossing, thereby communicating this information directly into the vehicle were it can be observed without relying on external signs or indicators which could be obscured from the operator of the vehicle. How best to use the newer technologies is still being debated as well as how much these systems will cost and what business model will support the development and maintenance of these systems.

With the adoption of this new technology, questions arise as to whether vehicle operators can handle additional information and how it should be presented to them. Human factors issues will have to be addressed in order



PROCEEDINGS 138/208

for such a system to be seamlessly integrated into the driver’s normal routine of managing dashboard indicators. There are many human factors implications of such systems that need to be addressed. What types of messages are adequate to provide to the operator, when and how often? Is an acknowledgement by the operator necessary? Since current systems sometimes provide both an advanced warning (typically signage) to a level crossing and an active system providing flashing lights and gate activation based on train approach, is it necessary or prudent to provide a train approaching warning to the operator inside the vehicle? Some feel this would provide a motivation for the operator to accelerate in order to beat the train, but it is clear that this type of violation occurs with current warning systems in the United States.

Introduction

Level crossing incidents continue to be a problem area for rail operators and motorists. With today’s faster trains, increasing freight rail traffic and continual increases in highway traffic, the problem will persist. Add to this increased activity the larger number of distractions within vehicles today, such as the widespread use of cell phones and entertainment systems, incidents will continue to occur. In-vehicle warning systems are necessary to provide motorists with clearer notification of grade crossings in a manner that will overcome the increasing numbers of driver distractions within a motor vehicle. The U.S. Department of Transportation (USDOT), Federal Railroad Administration (FRA) is addressing the situation by consolidating and closing as many crossings as possible where practical and providing educational emphasis to motor vehicle operators. The FRA has also provided research funding for the development of enhanced warning devices and safety treatments at level crossings, such as four quadrant gate systems, and has worked with other Department of Transportation modes to improve awareness near level crossings.

The USDOT Federal Transit Administration (FTA) as part of its goal to increase the use of public transportation on transit facilities is opening new level crossings and is proactively focusing on the use of new technologies such as Intelligent Transportation Systems (ITS) to enhance the safety of level crossings. FTA and FRA have taken part in activities associated with the USDOT ITS Joint Program Office initiatives that include safety at level crossings and intersections in general. Many of the functions established in the initial designs for Cooperative Intersection Collision Avoidance System (CICAS) and Vehicle Infrastructure Integration (VII) programs can be applied to level crossings.

Warning treatments to date have traditionally consisted of a combination of signs, bells, flashing lights, and gates as shown in Figure 1. While these approaches have assisted in reducing the number of incidents at level crossings, they do not provide a proactive approach to issuing a caution alert that can overcome some of the in-vehicle driver distractions. There have been a number of prototype systems that were developed and demonstrated to provide in-vehicle warning systems to motorists. These attempts have used custom components and non-standard technologies with varying levels of effectiveness. While some of these systems have shown that in-vehicle systems can provide safety benefits, a thorough human factors analysis has not been part of any of these projects. It is important to note that there has not been a thorough study to determine the effectiveness of the system after it has been in operation for a substantial period of time.



PROCEEDINGS 139/208

Figure 7: Level Crossing

Proposed In-Vehicle Methodologies When contemplating the implementation of in-vehicle warning systems at level crossings, two concepts appear to make the most sense based on the technology advances that have occurred [1]. Either approach is capable of providing the mechanism to assist with in-vehicle messaging to operators.

Method 1: In-Vehicle Advisory via Wireless Communication with Roadside Device

The first concept is based on the use of dedicated short range communication (DSRC), a 5.9 Ghz Radio Frequency system used for short range data transfer. DSRC is being used as the technology for functions such as electronic toll collection (ETC) as well as the communication system for Cooperative Intersection Collision Avoidance System (CICAS) and Vehicle Infrastructure Integration (VII) programs at the USDOT Federal Highway Administration (FHWA). It is the planned implementation of this technology within the VII and CICAS programs that make the inclusion of in-vehicle warnings at level crossings seem a logical next step.

Since vehicles will be equipped with DSRC receivers for purposes of collision avoidance and stop sign warnings, vehicles could also be adapted to receive “Approaching Level Crossing” or “Train Approaching” warnings. The initial focus could be to mimic the same functions as a stop sign warning whereby the vehicle operator is alerted to the fact that they are nearing a level crossing and extra caution should be taken. Level crossings can be retrofitted with the necessary roadside DSRC equipment to initiate the warning process similar to the device that stop signs will be outfitted with. Figure 2 shows what past prototype systems have used for issuing warnings within a vehicle as well as a typical navigation display used today.

There may be a desire to equip selective level crossings with this functionality. Perhaps only those crossings that are considered high risk based on past history, configuration, or frequency of use would be equipped.

Method 2: In-Vehicle Advisory via GPS Location Navigation System

The second concept that could provide cautionary notification of approaching a level crossing is based on the use of today’s widely available navigation systems. Many systems are offered as standard equipment in vehicles as well as the widely popular after market add-on equipment. Combined with precision global positioning satellite systems, these navigation systems could play a key role in providing increased safety at level crossings. Current systems provide detailed location and direction information based on GPS location to operators either via detailed maps displayed on system display units or through the use of voice generated driving instructions. Detailed maps are stored within these systems. The mapping data is frequently updated



PROCEEDINGS 140/208

and the system allows for addition of points of interest. These systems rely on signal availability from GPS satellites as shown in Figure 3.

Since many, if not all, navigation systems allow for updates to their map database as well as provide for optional points of interest, it is suggested that the location of all or at a minimum all high risk level crossings could be added to the navigation systems databases. This could be done through the addition of precise latitude and longitude points for level crossings into the base maps that are provided with navigation systems or through an add on feature that many units provide for updating of maps or for inclusion of points of interest.

This effort could be coordinated with such activities of centralizing all of the level crossings in the U.S., for example into a national database which provides not only precise location of level crossings, but also details of ownership, type of warning treatments at the crossing, street information, etc.

Again the purpose of having such a system is to provide a simple caution advisory to the vehicle operator, alerting the operator to the level crossing that they are approaching. A distance to function could determine when to provide the advisory to the operator. If the onboard database of the crossings is capable of containing additional information, there would be an opportunity to provide potential critical information for special case crossings. An example of this is to provide an additional caution of an unusual crossing configuration that may provide higher risk for certain types of vehicles such as low-clearance trailers, school buses, agricultural equipment or construction equipment.

Figure 8: (L) Prototype in-vehicle warning system ( R) modern navigation system

Integrating an in-vehicle warning system with the railroad system may allow the implementation of providing advance messaging advising the motorist not only of the time of train arrival to the level crossing, but also if there will be an extended delay at the crossing. Integration of this advanced system with a mapping program could also provide alternative routing information to the operator thereby relieving congestion in the area.

The potential for implementing any of these concepts alone or in combination with one another is real and exists today [3]. The technology is commercially available and there are standards that already have been developed to address the interface between the highway and rail subsystems [4]. The critical step to implementing such a system is developing a funding mechanism to develop a prototype test bed that can be used to demonstrate the concepts and further analyze the human factors issues that are apparent in providing warnings to vehicle operators.

ITS have been introduced to improve the operating efficiency and safety of rail transit and highway systems. Many concepts have been explored, prototyped and demonstrated at various conferences and expositions. It is extremely important that some of these systems be deployed within the nation’s infrastructure to demonstrate that they can actually provide the services for which they have been developed. Recent initiatives such as SafeTrip-21 is taking a giant step in that direction by focusing in on those technologies that can be introduced and provide immediate benefit to the travelling public.

In-vehicle warning systems can provide some quick benefits related to level crossings. Even though great improvement has been accomplished in increasing the safety of crossings through other methods, there are still many incidents that could potentially be avoided by providing advanced warnings to motor vehicle operators through the use of in-vehicle warning systems. Since current FHWA projects include a function to provide STOP sign warnings to operators, it is along these lines that an operator could also be cautioned to the crossing that they are approaching. Since a direct tie in to a train control and/or crossing controller introduces more complexity to the system, it makes sense to adopt the independent stop sign concept for crossings as well.



PROCEEDINGS 141/208

Figure 9: GPS Communications

Human Factors Implications Previous prototype in-vehicle warning systems have attempted to capture the impact the systems have had on safety, but have not conducted detailed studies on how the drivers have reacted to these systems. The types of warning indicators used for these prototype systems was not based on a single approach, so therefore the results of each system were not easily compared and thus were not conclusive. Since driver distraction and information overload are two key issues with providing successful in-vehicle messaging systems, it is of the utmost importance that the development of these systems be conducted in an open manner that allows for a standards development process to occur resulting in a unified approach to providing in-vehicle warning systems. The human factors implications of such systems are fairly distinct. What types of messages are adequate to provide to the operator and when and how often? Is an acknowledgement by the operator necessary? Since current systems sometime provide both an initial approach warning to a level crossing and an active system providing red light and gate activation based on train approach, is it necessary or prudent to provide a train approaching warning to the operator inside the vehicle? Some feel this would provide a motivation for the operator to engage in a riskier maneuver and accelerate in order to beat the train, but it is clear that this type of violation occurs with current warning systems in the United States. There are several organizations including the Society of Automotive Engineers (SAE) and the International Organization for Standardization that continue to work on human factors standards for ITS [2]. Some of these standards provide guidance on the types of warnings to provide to the vehicle operator. The standards also discuss the format of messages, the content of such messages and how to physically convey them to the operator. Figure 4 shows just one of the many implementations for display of data to vehicle operators.



PROCEEDINGS 142/208

Figure 10: In-Vehicle Navigation System

What makes the idea of in-vehicle warnings attractive is the fact that previous research on the habits of vehicle operators suggests that they do not provide a stepped up readiness when approaching a crossing during their normal travelling activities. With the implementation of some type of in-vehicle warning system that is commonplace, it is hoped that this will provide the notification and awareness needed to engage the operator resulting in less risky behavior and increased safety. Outlook Only after further testing and understanding of the prototype systems that have been developed thus far will there be an opportunity to begin a rollout of these systems into everyday use. It is important for the long-term success of in-vehicle systems that they are developed in-concert with the overall initiatives being managed by the ITS Joint Program Office. A successful system is one that has been integrated and tested as a system in whole and has undergone a thorough human factors analysis to make sure the operators and everyday users of the system understand and benefit from its functionality. A recent refocusing effort being managed by USDOT’s Research and Innovative Technology Administration (RITA) called SafeTrip-21 is aiming to bring together research that has been completed and implement those that are most promising of providing benefits today through the use of test sites. The objectives of the test sites are to speed up the development of the USDOT’s VII initiative and enhance what research has been accomplished in the areas of electronic information, navigation, and communications technologies. This will further enable the nation to accomplish several goals in the area of transportation including: reduction in motor vehicle crashes, alleviating traffic congestion, increase use of public transit and ride sharing as well as promoting motor freight efficiency and safety. The SafeTrip-21 initiative will continue to explore the feasibility of deploying this technology nationwide. Several states and localities along with automakers and equipment suppliers will continue to share lessons learned from previous and ongoing operational testing. In comparing the two approaches, neither one is technologically superior over the other, both have their advantages and disadvantages and both will require significant up front investment. The first method, implementing DSRC, lends itself to a truly integrated system since it will be built upon the foundations of the current, ongoing VII initiatives. The second method, utilizing GPS technologies, builds upon commercially available equipment and capabilities. It is through the further development of ITS systems and human factors analysis that one of the approaches will begin to rise to the top and provide the type of seamless functional safety aspects for ITS implementation. As further development of these types of technology continues, there may be other methodologies that will provide a superior solution to the problem and is a main reason why this type of research is encouraged.



PROCEEDINGS 143/208

Acknowledgements

We would like to thank the USDOT/FRA Office of Research and Development for their continued support and funding of this effort.

References

[1] Bousquet, P. E.; Peck, S. M. “Review of Intelligent Transportation Systems Applications at Highway-Rail Intersections in the United States, (2003).

[2] Huey, R. W.; Jenness, J. W.; Lerner, N. D.; Llaneras, R. E.; Singer, J. P. “Human Factors Guidance for

Intelligent Transportation Systems at the Highway-Rail Intersection”, (2006). [3] Peck, S. M.; Bousquet, P. E. “New Technologies in Intelligent Transportations Systems for Highway-Rail

Intersections”, (2003). [4] The Institute of Electrical and Electronics Engineers, “IEEE-Std 1570-2002, IEEE standard for the

Interface Between the Rail Subsystem and the Highway Subsystem at a Highway Rail Intersection”, (2002).



PROCEEDINGS 144/208


Risk assessment – is it a best possible method for LC categorisation?

Author(s): Witold Olpinski

Job Title: Senior specialist

Company: Centrum Naukowo-Techniczne Kolejnictwa (CNTK)

/Railway Scientific and Technical Centre/

Country: Poland

Resume of Speaker

Witold Olpiński received his Master of Science and Engineer’s degree from Electronics Faculty at Warsaw University of Technology in 1983. In 1998 he completed Postgraduate National Security Studies at the National Defence University. Since 1983 he has been working for the Railway Scientific and Technical Centre (CNTK) now as a Senior Research and Technical Specialist. He is a Co-ordinator of the Rail Transport Branch Contact Point of EU Research Programmes in CNTK. He represented Polish Railways in the Change Control Board and Control Command and Signalling TSI Working Group heading the subgroup for LC in the ETCS system. He is a member of the Technical Committee for Railways of the Polish Standardisation Committee. He participates in 6.FP “SELCAT” project and in European Level Crossing Research Forum (ELCRF) activity.

Abstract The paper is trying to give a positive answer to the question asked in the paper title. The substantial problem related to the harmonisation of the European LC, particularly from the point of view of the road user is considered. The variety of the LC equipment and different application rules (LC equipment behaviour, distances, timing etc.) whereas road signs are identical, exists throughout countries in the EU Member States and worldwide. It may cause lower awareness and vigilance of the road user and increased number of the LC accidents. The road infrastructure managers’ responsibility for the LC safety problem is underlined. The ERA basic LC types’ classification is discussed. The approach to a risk assessment as the basis of the LC categorisation is suggested in the paper. The advantages of such methodology are described. A step-by-step approach to the harmonisation of the LC protection equipment selection as the result of the LC classification is proposed in the paper. Introduction The international research society from several years has taken efforts to improve traffic safety situation on the railway level crossings (LC). This is the 10th time when the International Level Crossing Symposium is conducted. In spite of enthusiastic comments of some previous Symposiums’ participants, it is very hard to estimate the real impact of these events and their influence on the safety on the level crossings problem. Fortunately, the communities of professionals interested in the increase in the safety in this dangerous, troublesome area, where two different



PROCEEDINGS 145/208

kinds of traffic are crossing one another in one level are more and more active. One of possible and important ways is the research experience exchange. This activity manifests itself in coming into existence of the European Level Crossing Research Forum – ELCRF, the non-commercial, fully voluntary and still non-formal researchers’ organisation, actively supported by several participants with the leading role of the Rail Safety and Standards Board - RSSB from the United Kingdom. Also, the European Commission acceptance to subsidise the 6th Framework Programme project “SELCAT” – “Safer European Level Crossing Appraisal and Technology” is a good example, that problems in these troublesome spots seem to be very important. As most of the Symposium participants know, it follows and is connected with the third workshop of the SELCAT project. Level crossings are the subject of the European Railway Agency (ERA) activities, particularly from the point of view of the overall safety parameters of the railway traffic. Unfortunately, today a common approach appears, with which I absolutely disagree, that all accidents on level crossings are treated as purely railway accidents, due to the definition that such accidents include all those, where one of the collided vehicles is the railway vehicle. Even more, in that group often we may find all such cases, where the accident victim is a road user without any railway vehicle involvement, but the accident occurred inside the dangerous area of the LC, or even generally, inside the whole crossing area. I often publicly express my opinion that this approach should be immediately changed. Most accidents on the level crossings are caused by road users and they may be treated exactly the same as events when the car went off track and hit a tree or went downhill to a precipice. The only reason which should be taken into account is certain danger created by the irresponsible behaviour of the road user, which may cause a dangerous situation for the rail traffic, either for passengers or for goods transportation. Obviously, when the accident caused by a road user results in railway equipment damages, passenger casualties and transported goods loss, it should be treated as a railway accident, however, road traffic participant casualties and their material losses should not be treated and counted (in the traffic statistics) as the railway accident results. It would be much easier for me to accept the opposite option, when only such accidents should be considered as the railway traffic accidents, if they are caused either by the failure of the level crossing protection equipment (including mistakes of the manually operated level crossing personnel) or by unauthorised approach of the train to the level crossing area. All issues related to the level crossings, as relevant legislation, construction rules and methods, technical solutions, particularly warning and protection equipment, road signs, information and warning plates, their application, operation, sizes, shapes and colours etc. are probably the most diversified area in the whole road traffic around the world. These issues have not been included yet in Technical Specifications for Interoperability (TSIs) of the European Union railways because the only factor which may have influence on interoperability is the train “movement authority” limitation (or revoking) and possible speed restriction when the appropriate feedback (interlocking) exists and the level crossing protection equipment indicates that the level crossing is not protected, usually due to a certain failure. There are two main reasons which push me to express my personal point of view concerning the level crossings safety issue. These are as follows: the level crossing protection equipment development activities, which are absolutely not harmonised and the wide range of substantial differences regarding the overall circumstances of passing the dangerous area of crossing the railway line as it may be seen by the road user in particular countries, railways and even particular places. Due to the permanent increase in the international road traffic, it may be suspected that these differences may become the important factor decreasing safety increasing the potential risk for foreigners passing the level crossing in other countries than their own. Despite the fact that road signs are generally standardised, there are some national differences between them, but usually only regarding their size, colour and some jointly used additional information. Unfortunately, application of the road signs that warn road users about approaching the level crossings is probably most diverse among all of standardised road signs. The road user may find vital differences between distances (even twofold) from warning tables to the level crossing, different warning sign (either with the steam engine or with the fence on the drawing) while there is nearly the same warning and protection equipment, depend on the national regulations. Moreover, there are fundamental differences among warning times, for example from the beginning of the warning to lowering of the barriers or from activation of the level crossing equipment (by the approaching train) to reaching the LC area by this train. There is also a wide range of light signals to warn road users, including among others specific solutions: single and double red lights, permanent or flashing, with additional



PROCEEDINGS 146/208

white light (active when it is possible to pass the LC safely) additional lamps, arrows, etc. to inform about the second train approaching (on the other track of double or multiple track line). The variety of sound warning devices and their operation rules is also great. LC types classification It is necessary to stress the basic problems that we encounter analysing the level crossings subject. The long list of problems starts with the differences in terminology used. One of factors taken into account during analysis of safety on level crossings is the type of the technical equipment used for warning and protection the level crossing area, mainly for the road users, to protect them while the train approaches, however, sometimes also to inform the train driver that the level crossing area is protected and safe. In the beginning, the LC classification in some countries is called as the categorisation of level crossings, while in others we are talking only about different level crossing types. Generally I notice that the level crossings classification differs from country to country and from one assembly to another and it is based on three specific approaches: - technical approach, where the level crossing type/category is related to the set of technical equipment

used for warning and protection purposes, starting from not equipped LCs, one equipped only with permanent road user warning signs (as the St. Andrew’s Cross, which, by the way, also has a variety of sizes, shapes and colours), LCs with light and acoustic warning signals, up to the LCs equipped with half barriers and full barriers and with a wide range of different devices, which may be used in a certain LC equipment type;

- functional approach, where the basic division may be done among them to the passive and active LCs, where “passive” means that the warning and/or protection installed, if any, remains unchanged independently of the rail (and road) traffic situation and “active” LCs group includes all those, where the warning and/or protection function is activated while the train approaches the LC area; the active group is then subdivided to manually and automatically activated; this approach, is used by ERA for the “basic types” classification, which will be discussed later;

- application approach, where the LC classification is based on the traffic requirements to ensure the certain protection level depending on rail and road traffic density, maximum allowed train speed, visibility conditions etc.

As usual, in a number of particular cases a hybrid classification is used. In practice, the brief phrase “level crossing type” is often identified with the level crossing protection equipment type used on the particular level crossing. ERA basic LC classification Aiming to introduce the basic order in the general level crossing safety problem related activities, the ERA proposed the LC types’ classification, as shown on Figure 1. This classification will be called further as the basic one.



PROCEEDINGS 147/208

Figure 1. The ERA classification of the LC types. In the first step, the ERA basic classification distinguishes between active (group A) and passive (group B) level crossings. The simplest explanation of the passive type of the LC may be such, that in this group we combine all level crossings equipped with any warning signs, plates, devices or any other protection equipment, which state is permanent and totally not dependent on any traffic situation. This group includes also non-protected level crossings. In opposite, the level crossings protection equipment on any active level crossing reacts somehow by changing the state (warnings and/or protection) when the train approaches to the level crossing area. In the second step of the ERA basic classification, the main difference between two subgroups of the active level crossings is the activation method of the warning and protection equipment on the level crossing. If the process, starting from the approaching train detection up to the state change of the protection and warning equipment happens without any manual intervention (by the level crossing keeper or any other personnel), the group is called as the “Automatic protection/warning LC” (group A.1). If anywhere in the process, the man activity is needed (usually to operate locally the position of barriers), such level crossing is classified to the group called “Manual protection/warning” (ERA group A.2). Each of these two groups is then divided into three identical subgroups, depend on the road side equipment used: road side protection, as barriers, gates (groups A.1.1 and A.2.1, respectively), road side warning – optical, acoustic, physical (groups A.1.2 and A.2.2) and road side protection and warning (group A.1.3 and A.2.3). It is a very general and relatively simple classification, however, the theoretical completeness and symmetry of the structure is not fully compatible with the practical solutions. In real installations, it is rather impossible to find, for example, the automatically operated LC with road side protection (as barriers) and without warning lights the same time. So, the group A.1.1 is empty in practice. A similar situation occurs with groups A.2.1 and A.2.2. This classification may be used only as the first step to unify the reporting rules for the EU Member States and to obtain the general overview of the safety parameters of the particular LC type groups. This classification will be used by the ERA as the element of the Common Safety Indicators in order to estimate the required Common Safety Targets.



PROCEEDINGS 148/208

Human factor issues Due to the lack of harmonisation of the level crossing protection equipment development, in several countries and laboratories there is undergoing continuous work on new technical solutions. The main basic development direction is focused on the various methods of the obstacle detection in the level crossing dangerous area. The result of such detection is the influence on the train movement using various signalling and communication means and methods, depending on the different train speed and distance from the level crossing. Also new road side level crossing users warning and information equipment solutions are being developed. Several independent, not harmonised activities on the road users’ awareness increase are also being performed in particular areas and countries. It is important to stress, that some solutions appropriate in one place (region) may be absolutely not effective in another area due to different local habits, customs, regulations and law. For example, the very good solution to physically protect railway traffic from unsafe road users behaviour – the concrete flaps of the road surface moving up in front of the level crossing – effective in Russia, probably will be never accepted in most other European countries as the possible reason of road users injuries. In order to drive a car it is necessary to obtain a driving licence. However, there are different rules, requirements and tests in particular countries. In spite of this, with some exceptions, the driving licence is generally binding. Typically, only professional car divers take specific predisposition tests. Road users, including all car type drivers and pedestrians represent all possible types of personalities determining their behaviour while passing the level crossing. I suppose that the level crossing protection equipment available is usually selected and installed appropriately using its features. In my opinion, partly based on my personal experience, there is a certain natural awareness and reaction border. Passing below that border, which may result in the accident, may be sometimes caused by some physiological reasons. Taking into account the overall population of road users using the level crossing, there is a certain probability of accident, really independent on the protection measures used. This conclusion is not denying the reasonability of introduction all available LC protection measures which are taking into account a human factor. The very good example of the various circumstances related to human factor issues consideration in the protection measures selection process is the software tool developed by the Rail Safety and Standards Board. Although the tool is oriented to the UK road and railway system, it is a very good example of the computerised support of the level crossing protection equipment and the other measures selection taking into account the accident risk caused by the human factors. The tool provides guidance on appropriate risk reduction measures. It allows developing the level crossing protection project appropriate for the certain human factor related to risk local circumstances by building it from bricks of solutions available “on the shelf” of the computer database. The approach used in the RSSB tool is similar to the idea of the final, structural approach, which I need to present in this paper. It shall consist of the series of subsequent phases. The first one shall result in the common definition of the LC classes. Risk assessment based LC protection equipment class ification Let us introduce the term “class” to incorporate all LC protection equipment types (LC types), which all behave the same from the road user’s point of view. It means that all LC types in one class have exactly identical timing, the set of warning and protection devices as well as all road signs used including their shapes, sizes, possibly colours and exact rules of their application. What is the most important, this classification should be based on the road user’s side view, i.e. all, may be various solutions regarding the technology or additional subsystems applied, shall be seen the same by the road user for the whole particular LC class. Each particular class shall guarantee a certain, minimal protection level. The classification shall not reduce the possible innovation and equipping the certain level crossing equipment type with any new subsystem which increases the protection, however, in the particular defined class, any system development should not influence the warning and protection equipment behaviour as visible for the road user. Obviously, it is possible and even recommended, that several classes ensuring different minimal protection level will be not distinguishable by the road used. First of all, there should be decided how many different level crossing equipment classes will be used on the whole railway network. I suppose that the reasonable number of different LC equipment classes shall include



PROCEEDINGS 149/208

from 5 to 9 guaranteed protection levels, however, the number of different level crossing types visible for the road user shall be reduced probably to not more than four. The second and probably the most important step in the structural approach proposed here shall unify the method used for the selection of the required protection level for the particular level crossing. The current situation varies substantially from country to country. The introduction of common rules usually will require changes of the local regulations, sometimes also in the national legislation. So it is possible to imagine, that the first area, if applied anywhere, would be possible in the European Union. Today particular countries, and sometimes even particular railways (their Infrastructure Managers) or administrative areas use different rules to organise safe railway and road traffic on level crossings. The wide range of rules and methods to define certain safety protection requirements for the particular level crossing is applied. It starts from very simple rules, as the calculation of the “traffic product”, which combines the volume of rail and road traffic on the basis of the traffic flow measurements performed in the well-defined manner. The achieved result is translated to the certain LC type required. For example, such an approach is defined in the current regulations being in force in Poland. This method is relatively simple and effective, however, it has several weak points. First of all, the “traffic product” may be exactly the same in each of three very importantly different situations from the traffic safety point of view, i.e. when: - the railway traffic is very low and the road traffic is relatively dense; - the railway and road traffic flows are on average; - the railway traffic is high, but the road traffic there is relatively low. In each of the mentioned cases, the accident danger is different due to such reasons, as for example the different time of road traffic breaks required separating both traffic flows. In fact, much higher probability of the accident emerges in case of the bigger railway traffic density. The set of important weak points of such deterministic method as the “traffic product” calculation includes for example the following: - the random structure of the traffic flows, particularly in the case of the road traffic; - the length of the road vehicles queue awaiting for the rail vehicle pass through the level crossing area

and the influence of these queues for the traffic in the neighbouring area (particularly in cities); - the impact of the particular objects in the service area of the road passing through the level crossing on

the volume and the distribution in time of the road traffic; - the rail and road infrastructure state in the level crossing area, including distances from the level

crossing, which has influence on the traffic through the level crossing. The road traffic in most European Union regions, particularly in the new member states, increases relatively fast. The deterministic method, based mainly on the traffic flow measurements is probably outdated. Thus, the second step suggested in this paper should result as a target in the introduction of the probabilistic method of the certain safety protection requirements definition for the particular level crossing, possibly and eventually in all countries. This method is generally called as the risk assessment of the accident occurrence on the certain level crossing. The traffic flow across the level crossing is significantly different during the separate time intervals, for example in the separate 2-hours portions of each twenty-four hours period (day). Instead of the simple traffic volume measurement, the traffic model should be built for the particular level crossing taking into account the results of traffic flow measurements in the 2-hours (or other duration) time intervals. The LC traffic model shall allow for estimation of the peak-hour traffic density and its relation to the average daily traffic volume, simulation of the road traffic in the neighbouring area and the accident risk assessment. The achieved results will be the basis of the LC protection equipment class selection, however, they shall not be directly applied for that purpose. In each particular case, the results obtained using the traffic simulation on the certain level crossing shall be analysed and possibly corrected by the appropriate panel of experts. The expert group shall be selected using the criteria well-defined in the relevant regulations. The experts’ panel shall consist of:



PROCEEDINGS 150/208

- Railway authority (usually the infrastructure manager) representatives; - Road authority representatives; - Police (usually responsible for road traffic issues); - Necessarily, local authority representatives; - Other experts depending on the local geographical and economic structure. The experts’ panel correction shall consider the local circumstances having influence on the possible accident risk. Such circumstances shall include the location of such objects generating the traffic flow on the level crossing as for example: schools, hospitals, big industrial plants, commercial centres, sport facilities etc. The possible range of accident risk values, i.e. the probability of the accident, shall be divided into several subsequent intervals, which number should be the same as previously defined number of the LC equipment classes. In the intermediate step, it shall be defined the direct relation between the accident risk assessment result (the certain accident probability interval) and the particular LC class. It should not be forgotten that one of the most important factor which has the direct influence on the number of accidents is the road users’ awareness. All available measures shall be applied to increase it. One of possible methods is a wide advertisement campaign. It shall include all well-known advertisement techniques. Such a European-wide (or worldwide) campaign shall be the necessary element of the structured approach implementation to the level crossing protection equipment type selection on the basis of the accident risk assessment. Conclusions In conclusion, the paper presents a structured approach to the issue of supplying level crossings with warning and protection equipment which should be as a target applied widely all over the world as soon as possible to reduce the number of level crossing accidents taking also into consideration people’s increasing international mobility. The level crossing warning and protection equipment, called briefly the level crossing type (or class) shall be selected on the basis of the unified approach, by the accident risk assessment. The minimal protection level of the particular LC class shall directly correspond to the estimated accident risk there. The overall level crossing warning and protection equipment application and behaviour, as seen from the road user point of view, shall be identical for at least one from the possible LC types and it shall be unified eventually for all countries, starting possibly from the whole European Union.



PROCEEDINGS 151/208

Session 8 - Engineering and Operation Chairman: Yves Mortureux, UIC/SNCF, France

The impact of changing legal powers for public road level crossing regulation in Great Britain, and use to in fluence better partnership working.

Author(s): Andrew Harvey Job Title: HM Inspector of Railways

Company: HM Railway Inspectorate, Office of Rail Regulation Country: United Kingdom

Resume of Speaker

Andrew Harvey graduated in Electrical Engineering from Bristol University in 1972, and trained with British Railways Western Region in Signal and Telecommunications Engineering. He worked in a series of positions with a signalling equipment bias and a significant involvement with level crossing operation and incident investigation. He also served in the Territorial Army (Royal Engineers) for 12 years. In 1988 he took a career move to join the Department for Transport to work on motorway road signalling and communications. He transferred within the Civil Service in 1992 to join HM Railway Inspectorate in the Health and Safety Executive (the railway safety regulator) as a Field Inspector, and subsequently moved as an “expert” to the Inspectorate’s Level Crossing National Expertise Team. He has developed a special interest in the complex area of level crossing legislation. He is a Chartered Electrical Engineer, and a Member of the Institution of Railway Signal Engineers and of the Institution of Electrical Engineers.

Abstract

The paper considers the changing legal framework for level crossing protection in Great Britain and how this is currently being used to support better road/rail partnership working to improve level crossing safety. The paper outlines the origins of level crossing legislation in Great Britain, and looks at the recent changes which have enabled level crossing protection requirements to be placed additionally on local traffic authorities rather than just the railway operator. Revised format level crossing Orders being trialled assist those in control of the most significant risk at level crossings – the local traffic authority – to take ownership of the road related problems, but permits this to be put in perspective of overall road safety. This is seen as a first step within the existing somewhat complicated and out-dated legal framework. Further measures may be needed to reinforce partnership working between road and railway authorities and production of agreed longer term joint strategies, which should result in use of available (public) funds to secure best overall results for society as a whole, rather than providing costly increased engineering protection to an often poor compromise between road and railway needs.



PROCEEDINGS 152/208

Introduction The content of this paper reflects the views of the author and should not be taken necessarily to represent the views of the Office of Rail Regulation. The powers for creation and legal requirements for protection of road level crossings on railways in Great Britain originate in the legislation authorising the original construction of each particular piece of the railway network, but there are also links to road, rights of way and safety legislation. Since railways were first built, level crossings have required adapting to meet the changing needs and demands of road and railway users. This paper records some of the principal developments in legislation relating to crossing protection and illustrates how the current legislation is being used to improve level crossing safety and encourage partnership working between road and rail authorities. It also records the development of proposals to review and possibly update or revise the legislation relating to level crossings. Historical Background – Level Crossing requirements in authorising Acts The construction of railways in Great Britain began to become significant in the early nineteenth century (although some railways pre date this) and the principal means of authorising and regulating the construction and use of railways was by individual Acts of Parliament. As a railway was seen as presenting a significant risk to the public and occupiers of land through which it passed, there was a general requirement for railways to be fenced throughout their length. In 1835 a general Act [1] was passed relating to highways, which included requirements where railways crossed public carriage roads on the level. This was the first general piece of legislation to regulate level crossings, and it required gates to be provided (normally closed across the ends of the highway), and “good and proper” persons to operate them to prevent danger or damage to users of the road. The requirements were extended to include turnpike roads, and also highways in Scotland in an amending Act [2] in 1839. In essence these gates, when closed to the road, served as a continuation of the railway fence, and when open to the road, fenced off the railway. The first alteration to these requirements was made in 1842 [3] which gave a power to permit the normal position of the gates to be varied from the position laid down in the original Act, where required and authorised. A power was also created to permit railway companies, where required for public safety, to replace, at their own expense, any level crossing with a bridge or underpass where authorised, subject to any appropriate compensation to parties who were affected. These general requirements were further expanded for new railways after 1845 in further Acts [4,5] which were passed, in the interests of standardisation and of economy of parliamentary consideration, to set out standard clauses to be incorporated in each subsequent railway Act. Because of the differing legal system in Scotland, a separate Act [5] was passed for application there. Both Acts applied a restriction of railway speed to 4 mph at any road level crossing adjacent to a station. These Acts are noteworthy in that they required that any public carriage roads or turnpike roads should normally be crossed by an over or under bridge unless specifically and individually authorised to cross on the level in the individual railway Act. For other highways or roads, authorisation for them to cross the railway on the level had to be obtained in the form of consent of two or more justices (or the sheriff in Scotland). Thus, at the time of construction of the railway any public level crossing had to receive some form of official authorisation. These Acts, where a level crossing was authorised, made different provisions for public carriage and turnpike roads (railway operated gates as in the 1835 Act normally closed across the road), other public highways (suitable user operated self closing gates or stiles) and private roads (outward opening user worked gates). However there was no explicit provision to cater for changes in status of the road after the railway had been constructed. Further general requirements were added, for subsequent railway Acts incorporating them, in an Act in 1863 [6] which additionally required each public carriage road to be provided with accommodation (a lodge) for the crossing attendant and required trains not to be shunted or stand over such crossings. A power was also added permitting where necessary for public safety, the railway to be required, at its own cost, to replace any public



PROCEEDINGS 153/208

carriage road level crossing with a bridge, or carry out other works to remove or diminish the danger arising. No evidence has been found of the recent exercise of these powers. The requirements of these Victorian statutes remain in force today – but use often outdated terms (such as turnpike roads) and are less easy to understand than modern legislation. Provision was made in an Act in 1933 [8] that the normal position of the gates at public carriage road crossings on any railway could be varied by a Direction made by the Secretary of State if requested. It also repealed the 4 mph speed restriction at road crossings adjacent to stations. Light Railways Some railways which were proposed were referred to as “light railways” usually local in nature and using more flexible engineering standards and often of a lower line speed. In 1896 an Act [7] was passed which permitted the construction of light railways to be authorised by Light Railway Commissioners using a statutory Order, rather than requiring an individual Act. These Light Railway Orders usually incorporated relevant clauses from the earlier Acts as appropriate, and included sometimes more relaxed requirements for individual level crossings appropriate to the nature of the line. Modernisation In 1948 the national railway network was nationalised, and came under the control of the British Transport Commission, which reported to the Minister of Transport. In common with industry generally following the Second World War, material and labour costs were significantly higher, and the costs of maintenance and supervision of crossings were found to be greatly increased. Delays to the increased levels of road traffic at some level crossings were also becoming unacceptable. In 1954 powers were given in an Act [9] to enable the Minister of Transport on request to consent to the use of lifting barriers instead of gates at any public carriage road level crossing, but retained the need for local supervision. No provision was made in this legislation for the use of traffic signs or signals. A joint report was prepared by the Ministry of Transport and the British Transport Commission in 1957 [10] which proposed the adoption of established continental practice with the further use of lifting barriers instead of gates, remote operation of crossings and use of automatic half barrier crossings. This required further legal powers to enable the requirements in the original railway Act to be varied, and to enable associated road traffic signs and signals to be placed on the road in accordance with the requirements of more modern road traffic legislation. These powers were authorised in an Act in 1957 [11], which gave the Minister power to make an Order on request of the railway operator to vary the protective arrangements to be provided at any public carriage road level crossing on the national network. These powers were extended in 1968 [12] to include other road crossings to which the public had access. In 1983 these powers were updated and re-enacted in dedicated legislation [13], and the scope increased to cover any railway rather than just the nationalised network, reflecting the needs of operators of lines no longer part of the national network subsequently taken over or re-opened by “heritage” operators and others. The Level Crossings Act also extended the scope of the Order making process to include any highway or other road to which the public has access. The Act gave very wide powers (such provision as the Secretary of State considers necessary or expedient for the safety or convenience of those using the crossing) and any Order automatically disapplies any earlier conflicting requirements. The Act also automatically ensured that traffic signs and signals provided in compliance with the Order also complied with road traffic legislation for legal placement and effect. Other legal provisions relating to level crossings were made where changes were made to level crossing status (eg private to public), or where a new road was required to cross the railway on the level. These were often catered for by individual clauses in one of the private Acts of Parliament sponsored by the British Railways Board about once a year. These Acts normally empowered the Secretary of State to consent to protective arrangements other than gates at particular level crossings. One particular Act [14] extended the requirements



PROCEEDINGS 154/208

of the 1845 acts relating to accommodation crossings to crossings on parts of the railway network authorised by pre-1845 Acts.

Revised powers for authorisation of railways The Parliamentary Joint Committee on Private Bill Procedure reported in 1988 [15] and recommended changes to the process whereby railway and other similar enactments were scrutinized. The recommendations of the report were accepted, and In 1992 an Act [16] was passed which enabled railway works in England and Wales to be authorised through an Order process, rather than requiring an Act, in a similar fashion to light railway Orders (which it superseded and repealed in relation to England and Wales after a transition period). These powers under the Light Railways Act were used under these transition arrangements, to transfer the rights and obligations of a part of the national railway network to another undertaking, and to permit “new” railways to be opened along the alignment of former railways. These more recent Light Railway Orders often included requirements for modernised level crossings appropriate at the time of transfer. Similarly some Transport and Works Act Orders contained provisions relating to modernised level crossing protection. Similar Transport and Works Act Order making powers have recently been brought into effect in Scotland through an Act in 2007 [17] Scrutiny of any proposed Transport and Works Act Order was undertaken through a public inquiry type process rather than through the parliamentary Bill process The previous “model clauses” in earlier legislation were generally incorporated, and further “model clauses” were defined. Whilst the authorisation process was simplified (by using an Order process rather than an Act), as far as level crossings are concerned, it still in general applied the same requirements as earlier legislation. This Act also made provisions governing the use and protection of “private” crossings which fell outside road traffic legislation – those on roads to which the public did not have access. The Act also gave a power to make Regulations concerning the types of traffic signs at such crossings. This power was subsequently exercised in Regulations made in 1996 [18] and these Regulations specify the signs to be used at footpath, user worked and similar crossings

Enforcement Up until this stage, the requirements in legislation (and subsequent amending Orders) were broadly prescriptive, and any failure to meet the requirements could only be formally addressed in some cases by the “nuclear” option of prosecution of the railway operator. As the nationalised network was responsible to the Secretary of State, any failure to meet the requirements was normally satisfactorily addressed through correspondence. There was no facility for the Secretary of State to vary the requirements for a level crossing other than when an Order was requested. Where public usage of a level crossing changed significantly so that the protective arrangements were no longer fully effective, the railway operator had no obligation to propose any change, nor was there any power for a change to be required.

Railway reorganisation – and consequential changes in level crossing legislation enforcement

In 1992 a government decision was taken that the national railway network should be privatised [19], and in response in 1993 a paper [20] was prepared by the Health and Safety Commission for the Secretary of State for Transport making recommendations for the safety regime. As far as level crossings were concerned, much of the legislation referred to above was to be brought directly under the “umbrella” of the Health and Safety at Work etc Act 1974 [21]. This had two important effects. Firstly, the requirements laid down in most individual level crossing Orders would become directly enforceable by the safety regulator, and secondly, any required amendments for safety purposes to this level crossing legislation could be made through secondary legislation (Regulations) rather than requiring a further provision in an Act of Parliament. Direct enforcement of level crossing legislation through the Health and Safety at Work etc Act 1974 meant that there were a number of means to secure compliance with the law without resorting to prosecution. However



PROCEEDINGS 155/208

this enforcement could only be directed to those who were duty holders under the 1974 Act – principally the railway operator. Level Crossings Regulations 1997 These Regulations [22] were made under powers in the Health and Safety at Work etc Act, and repealed and amended some of the earlier Order making legislation now under the “umbrella”. They introduced a directly enforceable legal requirement for level crossing operators to comply with the requirements of a level crossing Order. The Regulations also amended the Level Crossings Act 1983 to permit the safety regulator to initiate an Order making process, and allowed the Secretary of State to make an Order even when the operator has not requested one. Content of level crossing Orders pre 2007 Orders made under the powers in s1 of the Level Crossings Act 1983 [13] may make such provision as the Secretary of State considers necessary or expedient for the safety or convenience of those using the crossing Level crossing Orders prior to 2007 were in a format that had evolved since the 1950s and available to railway operators as a template for the particular type of crossing. The Order itself contained a citation (the name of the Order), any revocation of earlier Orders, and date of coming into force. It required the level crossing operator to provide maintain and operate the protective equipment crossing in accordance with the accompanying schedules. The equipment and operation schedules varied considerably depending on the type of crossing to be provided, but typically included the following items:- Equipment Barriers – position, height, performance, markings and illumination

Road traffic signals (specified in road traffic legislation) Audible warning devices

Other road traffic signs (eg Keep Crossing Clear) Road markings – stop line and centre and edge of carriageway markings over the crossing and on the approaches Crossing telephones and lighting Railway signals and signs on the approach to the crossing Crossing operation equipment Cattle – cum – trespass guards Operation Road surface over the crossing How the crossing shall operate – timings etc. Sequence of operation for opening and closing the crossing Failure modes Operation of the railway over the crossing – interlocking, speed etc A number of items were in practice not included in the Order Warning road traffic signs on the crossing approach (a Highway Authority responsibility since crossings were first modernised) Vertical road profile on the approaches and over the crossing (automatic crossings) The Order making process was initiated by the railway operator, and included a statutory two month consultation period with the relevant planning and highway authorities. At the end of the consultation period, the Secretary of State would, if appropriate, make the Order, amended if necessary to take account of any representations received.



PROCEEDINGS 156/208

Areas of concern over safety and convenience issues As a result of ongoing monitoring and inspection of level crossing incidents and operation by HM Railway Inspectorate, several common general areas of concern became apparent which were not addressed by any provisions in level crossing Orders. The issue of lack of maintenance of vertical road profiles at automatic crossings was becoming increasingly apparent, and with it an increasing risk of a low vehicle becoming grounded on the crossing. The carrying out of joint periodic monitoring had lapsed, and the need for road and rail co-operation was not appreciated by the bodies responsible (both road and rail), which had often been subject to re-organisation. There was also some evidence that changes to railway line speed requiring additional cant to the tracks, or even routine mechanised track maintenance over the crossing were being carried out without considering the impact on the crossing profile. The road approaches/exits of some level crossings were subject to significant alteration without adequate consideration of the level crossing type and operation. Examples include the introduction of a one way traffic system on the road over the crossing, removal or suspension of relevant weight restrictions, construction of new road junctions adjacent to crossings, provision of road traffic management (such as speed humps, chicanes or road traffic signals), changes in land use, such as erection of new premises or houses affecting a traffic flow over the crossing. New road schemes were also found to have been designed without fully considering the impact on level crossing operation, or failing to capitalise on opportunities to close, or facilitate future level crossing closure. Minor road alterations and resurfacing work by local traffic authorities on the approaches to level crossings often resulted in the road markings required in the Order not being correctly replaced, and again there was confusion as to whether the matter was a railway operator or local traffic authority responsibility. General confusion was apparent on responsibility for maintaining road markings, tactile thresholds and traffic signs on the approach to crossings, and the responsibility for updating warning signs required by road traffic legislation. Instances were found of local traffic authorities permitting bus stops to be placed close to level crossings, and not applying road traffic restrictions to prevent parking in lay-bys provided for long slow road vehicles required to stop and telephone before using automatic crossings. Where road traffic levels were high, some local traffic authorities regularly expressed concern over apparently unduly long closure periods – although in some cases it was the increased traffic levels or changes in road layout which caused the crossing to be upgraded from an automatic half barrier to full barrier crossing with increased closure times. Information relating to road traffic accidents at level crossings was subject to different reporting arrangements from other road traffic accidents, and local traffic authorities generally were not made aware of problems at level crossings until contacted by the railway operator. There was also a significant concern that the legislative requirements relating to any particular level crossing were not easy to determine, often with requirements scattered through a level crossing Order and a number of amendments referring to out of date road traffic legislation, and different terms in the content and applicability of earlier Victorian legislation. The status of crossing (and original use) defined in the original Act had also often changed with times resulting in further lack of clarity. In some cases what were private roads had become public roads, and some bridleways (for use on horse and foot) redesignated as “byways open to all traffic” permitting additional use by vehicles. Amended powers in the Level Crossings Act 1983 An opportunity for a step change in vehicular level crossing management was brought about following consideration of level crossing Orders for the trial installation of median strips at three automatic level crossings. During the consultation on the draft level crossing Orders, it was found that there were difficulties with getting agreement with the relevant highway/local traffic authorities for the works at the crossing,



PROCEEDINGS 157/208

concerning the placement and maintenance of median strip equipment. Network Rail took an initiative to propose an amendment to the Road Safety Bill then being considered in Parliament, address this difficulty. The amendment proposed would amend the Level Crossings Act 1983 to explicitly permit requirements in level crossing Orders to be placed on local traffic authorities as well as the railway operator. There was some debate concerning who would fund any “additional” requirements, but the amendment was successfully taken forward by the Government and incorporated in the Road Safety Act 2006 [23]. Amended consultation process In progressing the amendments, opportunity was taken to update the consultation process in line with modern principles. The revised process incorporated a preliminary consultation by the railway operator with the safety regulator and the local traffic authority, which was then followed by the established process of draft Order circulation for consideration by the Secretary of State. In an ideal world this should result in any areas of disagreement being resolved in the first consultation round before the Secretary of State is requested to make an Order. Due to an oversight in the amendment of definitions, local planning authorities were excluded from the consultation process, but guidance still encourages their inclusion. Revised format for level crossing Orders Because of the areas of concern listed above, consideration was given to utilising these more explicit powers to improve level crossing safety. Whilst in theory it is for the level crossing operator to propose the contents of draft Orders, in practice, in the interests of standardisation, HM Railway Inspectorate has provided “template” Orders to assist operators. A decision was taken to update and extend the wording of the template Order, from the established pattern, to include additional schedules to specify equipment and duties to be carried out by the local traffic authority. Content of revised level crossing Orders It should be emphasised that none of these additional responsibilities is being imposed on local traffic authorities (or any other body) without consultation and consideration of any comments. Each requirement must pass the test of being “necessary or expedient for the safety or convenience of those using the crossing”. Consideration must also be given to the magnitude of the level crossing risk being addressed in balance with other road safety risks within the local traffic authority’s area of responsibility. One of the areas clarified in the revised Orders is the responsibility for the road markings on the road approaches, making the local traffic authority expressly responsible for all items on the approach to the road “stop” lines. Road traffic signs on the approaches are also included in the revised Orders, although these have always been a local traffic authority responsibility, albeit unrecorded. For automatic half barrier crossings, responsibilities are being placed on both railway operator and local traffic authority to maintain the vertical road profile within appropriate limits in co-operation with each other. A further duty is being placed on the railway operator to inspect and monitor all the items in the Order, and to draw the attention of the local traffic authority to matters which are their responsibility, such as visibility and legibility of road traffic signals and signs. Local traffic authorities in England and Wales have powers [24] to address vegetation overhanging a highway. Where level crossings are close to other signalised road traffic junctions, the need to interlink the road and rail systems to avoid conflict can be identified in the Order, and a requirement placed on the railway operator to provide appropriate controls and a defined interface point accessible to both organisations. An additional duty on the railway operator can be considered to provide information on incidents arising at the crossing, and for this to be provided to the local traffic authority for their consideration in the exercise of their functions in relation to the road.



PROCEEDINGS 158/208

A joint duty can be placed in the Order for the railway operator and the local traffic authority to agree a shared long term strategy for continued use of the crossing, and to consider, where appropriate, what measures might be necessary to permit partial or complete closure of the crossing. This can feed into the local transport authority local transport plan. Where local transport plans also feature increasing rail use and improved local rail transport, then the impact these have on increased level crossing closure time and road traffic delays can be evaluated, and a strategy developed based on society needs as a whole. Where road closure times are an issue, provision can be made in the Order to require the railway operator to monitor and minimise road closure times. A number of further “optional” paragraphs being included on a trial basis as appropriate for individual crossings:- Where there is a requirement for the local traffic authority to exercise its powers under other legislation (for instance imposing traffic restrictions such as a weight limit, or prohibition of parking in lay-bys for large vehicles), then the content of a paragraph in the Order needs to be carefully worded so as not to introduce conflict with other legal requirements – and importantly not mandate a local traffic authority to exercise powers which are discretionary or subject to further consultation. Where appropriate a requirement can be inserted in the Order for the local traffic authority to provide a contrastingly coloured anti-skid surface on the approaches to the crossing. At level crossings where the design of the crossing is based upon road traffic levels close to a threshold, then a requirement can be considered for traffic levels to be monitored, and for the local traffic authority to consult if there are proposals elsewhere on the road network which may impact this figure. In this sort of circumstances, if the local traffic authority becomes aware of proposals to close a parallel route which will divert additional traffic over the level crossing, then a duty could be specified to consult the railway operator in advance. A similar argument applies where an axleweight restriction has been placed on the road so as to ensure safe operation of the crossing, and a temporary suspension of that axleweight limit is agreed. Issues arising after circulation of revised Orders Objections have been received from two local traffic authorities that the warning signs on the approaches should be the railway operator’s responsibility, and that there is no budgetary provision for maintenance of these signs. As advance warning signs have always been the responsibility of the local traffic authority, it illustrates that the new Orders are assisting in clarifying responsibilities. Review of Level Crossing Legislation The use of these revised powers to make Orders gives the means to address, on a crossing by crossing basis, some of the immediate areas of concern. However the underlying complexity of applicable legal requirements remains. Requirements relating to a particular crossing can be scattered through different railway Acts, Directions and level crossing Orders with amendments. In 2007, the Department for Transport, supported by HM Railway Inspectorate within the Office of Rail Regulation, prepared a case for a project on level crossings to be included in the Law Commission's tenth programme for legislative reform. The case presented to the Law Commission was for a project to review the legislation that applies to level crossings in Great Britain and to look at the regulatory framework in which level crossings operate. At the time of writing this paper, the outcome of that application has not been determined, but an announcement may be made shortly.

Acknowledgements

Acknowledgement is made to colleagues in the Office of Rail Regulation and other organisations for their assistance in the preparation of this paper.



PROCEEDINGS 159/208

References

[1] Highways Act 1835 5&6 Will 4 C 50 [2] Highway (Railway Crossings) Act 1839 2&3 Vict C45 [3] Railway Regulation Act 1842 (s 9) 5&6 Vic c55t [4] Railways Clauses Consolidation Act 1845 8&9 Vict C20, [5] Railways Clauses Consolidation (Scotland) Act 1845 8&9 Vict C 33 [6] Railways Clauses Act 1863 26 & 27 Vict C 92 [7] Light Railways Act 1896 59&60 Vict C48 [8] s42 Road and Rail traffic Act 1933 23&24 Geo5 C 53 [9] Section 40 British Transport Commission Act 1954 2&3 Eliz II C55 [10] Level Crossing Protection – Report by officers of the Ministry of Transport and Civil Aviation and the British

Transport Commission 1957 HMSO [11] s 66 British Transport Commission Act 1957 5&6 Eliz II C33 [12] s 124, Transport Act 1968 C 73 [13] Level Crossings Act 1983 C 16 [14] Accommodation Level Crossings Act 1995 C8 [15] Report of the Joint Committee on Private Bill Procedure; 1987-88 HL Paper 97 HC 625 [16] Transport and Works Act 1992 C 42 [17] Transport and Works (Scotland) Act 2007 ASP 8 [18] Private Crossings (Signs and Barriers) Regulations 1996 SI 1786 [19] New Opportunities for the Railways, Secretary of State for Transport 1992 Cm 2012 ISBN 0-10-120122-2 [20] Ensuring Safety on Britain’s Railways, Health and Safety Commission January 1993 [21] Health and Safety at Work etc Act 1974 C 37 [22] The Level Crossings Regulations 1997 SI 487 [23] Road Safety Act 2006 C 49 [24] s154 Highways Act 1980 C66



PROCEEDINGS 160/208


Possible Measures for Safety Increase of ŽSR Level Crossings

Author(s): Ales Janota, Karol Rastocny, Jiří Zahradnik

Job Title: Vice-dean / University professor

Company: University of Žilina, Faculty of Electrical Engineering, Dept. of

Control & Information Systems

Country: Slovakia

Resume of Speaker

Mr. Janota was born in 1963. In 1986 he graduated from the Technical University of Transport and Communications (later renamed to University of Žilina, Czecho-Slovakia) in the study program “Railway Interlocking, Signalling and Communications”. Since 1986 he has worked with the university where he gradually got his academic degrees (PhD - 1998, Assistant Professor - 2003) and held different posts from assistant, senior lecturer, deputy-head of Department to vice-dean at the Faculty of Electrical Engineering. He is also a member of the Polish Academy of Science (Transport Commission, Katowice), Slovak standardization committee (TK 104 Control of industrial processes) and other bodies. His work deals with analysis and design of safety-critical systems, transport telematics and AI. He is an author of 2 monographs and ca 100 research and technical papers.

Abstract

The paper presents some of partial results obtained within the European 6FP research project SELCAT (No. TCA-CT-2006-031487), aimed at safety appraisal and technology of railway level crossings (LCs) seen in both European and the world-wide context. At the beginning of the paper some statistic data is presented to characterize the actual situation of using LCs in Slovakia, together with brief comments to its analysis. Despite the fact that statistic data from the observed period (2000-2006) identifies road users being responsible practically in all accidents we believe there is still a potential on the side of the railway operator to reduce a number of accidents at Slovak LCs. Therefore the attention of the authors is focused on proposing and presenting potentially usable measures that could increase safety of traffic operation at the LCs operated by the ŽSR (Slovak Railways). Separately there technical and organizational measures are discussed. Some of the proposed measures are specific for Slovak conditions only, however to a certain extent some findings can be generalized and possibly applied in other countries, too. Introduction Numbers of fatalities caused by accidents that occur at LCs represent only a minor part (ca about 1%) of all fatalities in the road sector. This percentage is much bigger for the railway sector; fatalities at LCs represent usually about 30% of all railway accident fatalities. This situation results in insufficient human danger awareness and lack of legislative regulations to stimulate road vehicle drivers’ discipline. Analyses prove that inadequate human behaviour is the main reason (in more than 90% cases) for the accident causes without



PROCEEDINGS 161/208

involvement of the safety system failure. Despite this, society labels most fatal accidents at LCs as a rail problem. Rail companies regard this as a particular problem because they cannot control the actions of road vehicle drivers and pedestrians at LCs. The following text shows some findings and results obtained by the authors participating in the SELCAT project and explaining the actual situation in Slovakia.

1. Statistic data about level crossings operated at Slovak Railways (ŽSR)

At present total length of Slovak railway lines operated by the ŽSR is 3661 km. There are 2351 level crossings functioning there, however only 46,8% of them is equipped with some kind of safety-related LC system (85,7% using relay-based technology, 3% electronic and 11,3% manually operated LCs). Generally, LC systems may be operated either with barriers or without them. LCs are operated at the lines with the maximum allowed line speed 120 km⋅h-1. The main corridor lines being just under reconstruction will allow the maximum speed 160 km⋅h-1. In Slovakia (unlike for example the Czech Republic), the use of LCs at corridor lines operated with the indicated maximum speed could be approved in exceptional and justified cases only. Table 1 summarizes statistic data related to numbers of LCs and their classes using the ERA classification. It is apparent that numbers given in the table have not been changed significantly between the given years. The essential changes will come into being due to reconstruction of corridor lines since many LCs will be cancelled and removed. 2000 2006 1 Total number of LCs, out of which: 2 500 2 351 2 - Passive LCs (equipped with St. Andrew’s Cross only) 1 385 1 251 3 - Active LCs (equipped with some kind of LC system), out of which: 1 115 1 100 4 • Active, operated manually by humans (with barriers) 154 124 5 • Active, operated automatically – without barriers 440 451 6 • Active, operated automatically – with barriers 521 525

Table 1: Numbers and classes of LCs at the ŽSR. Accidents occurring at LCs are usually closely watched by mass media despite the fact they represent only a minor part of all road accidents occurring in the Slovak Republic (see Table 2). These accidents are often linked with fatalities and serious injuries (see Table 3), therefore they should put our mind to their causes in order to propose such measures that (under acceptable economic conditions) make reduction of accidents at LCs possible. 2000 2001 2002 2003 2004 2005 2006 Passive LCs 35 30 32 17 27 35 22 Active LCs – manually operated 0 0 0 0 0 1 0 Active LCs – without barriers 47 20 30 14 26 27 33 Active LCs – with barriers 2 17 13 9 6 8 3 Total LCs 84 67 75 40 40 71 58 Total Roads 50 932 57 258 57 060 60 304 61 223 59 991 62 033

Table 2: Number of accidents at LCs and roads.

2000 2001 2002 2003 2004 2005 2006 Passive LCs 5 4 6 1 7 3 1 Active LCs – manually operated 0 0 0 0 0 0 0 Active LCs – without barriers 3 6 8 4 1 3 7 Active LCs – with barriers 1 3 0 0 2 1 4 Total LCs 9 13 14 5 10 7 12 Total Roads 628 614 610 645 603 560 579

Table 3: Number of fatalities at LCs and roads.



PROCEEDINGS 162/208

Tables 4 and 5 show accident distribution based on cause and the way of informing a road traffic participant about a state of the LC.

Cause Active automatic LCs with barriers

Year LC failure Human mistake

– railway staff Human mistake - road

traffic participant

Totally

2000 0/0 0/0 2/1 2/1 2001 0/0 0/0 17/3 17/3 2002 0/0 0/0 13/0 13/0 2003 0/0 0/0 9/0 9/0 2004 0/0 0/0 6/2 6/2 2005 0/0 0/0 8/1 8/1 A

ccid

ent e

vent

/ co

nseq

uenc

e [n

umbe

r/fa

talit

ies]

2006 0/0 0/0 3/4 3/4

Table 4: Analysis of accidents at active automatic LCs equipped with barriers.

Cause Active automatic

LCs without barriers Year

LC failure Human mistake – railway staff

Human mistake - road traffic participant

Totally

2000 0/0 0/0 47/3 47/3 2001 0/0 0/0 20/6 20/6 2002 0/0 0/0 30/8 30/8 2003 0/0 0/0 14/4 14/4 2004 0/0 0/0 26/1 26/1 2005 0/0 0/0 27/3 27/3 A

ccid

ent e

vent

/ co

nseq

uenc

e [n

umbe

r/fa

talit

ies]

2006 0/0 0/0 33/7 33/7

Table 5: Analysis of accidents at active automatic LCs without barriers. Analysis of statistic data has shown the following findings:

• Active LCs with manual warning/protection are the safest; on one side it is caused by permanent presence of human responsible for safety at the LC (which increases discipline of road traffic participants) and by existence of the mechanical protection, on the other side the significant fact is that this kind of LC is operated mostly at regional and low-intensity railway lines; however, despite positive effects on traffic safety such a solution is unacceptable from economic point of view and this kind of LC is step by step substituted with active LCs with automatic warning and/or protection;

• Practically all accidents caused at LCs operated by the ŽSR during the observed period (2000-2006)

were caused by road traffic participants;

• Number of accidents at LCs equipped with barriers is essentially lower than accidents at LCs without barriers (barriers represent a physical obstacle that can be less unseen than light signalling and non-respecting it usually requires more complicated manoeuvre of a road traffic participant).



PROCEEDINGS 163/208

Tables 1 up to 4 show statistic data from 2000 to 2006. Previous statistic data from the period 1996-1999 was published and is also available in [5]. This data was obtained from official statistics of the ŽSR and Slovak Road Administration.20

2. Statistic data in the context of European LCs

One of the results obtained in the SELCAT project was the rich database of statistic data provided by individual participants of the project. Comparison of safety of traffic process at LCs all over the Europe turned out to be a difficult task, especially for the following reasons:

• Initial ambiguous and non-uniform terminology or different understanding of the same terms (e.g. threat, danger, accident etc.);

• Non-uniform way of evaluation of accident consequences (different definitions for fatalities and their

evidence, serious injuries, slight injuries, material damages);

• Non-uniform classification or categorization of LCs and the way of their applications (for example in dependency on traffic intensity at the LC, on speed of transport means etc.);

• Differences in operation rules.

Just to illustrate the situation, Figures 1 and 2 present the graphs that can be used to compare number of accidents at LCs with different equipment in some of European countries (BG – Bulgaria, H – Hungary, PL – Poland, UK – United Kingdom, SK - Slovakia). Number of accidents is scaled – as a scale parameter we have used a number of LC systems of the given type applied in the country. The graph in Figure 2 does not contain data from 2006 for LCs in UK (unavailable). Considering the indicated problems with comparison, the graphs should be viewed with a grain of salt.

Figure 1: The illustration graph 1. Figure 2: The illustration graph 2.

On the base of more detailed analysis of available data we can state that from the viewpoint of safety at LCs Slovakia belongs to standard European countries.

3. Possibilities for increasing safety at LCs opera ted by the ŽSR

It is indisputable fact that “the best LC is no LC”, unfortunately building fly-over crossings is usually costly and long-time process. Traffic safety at railway LCs depends on two aspects: application of technical measures to reduce risk (building LC systems) and following the organizational measures, especially by road traffic participants. Procedures related to railway traffic operation at LCs of ŽSR are defined by the regulation [14]. Technical requirements for LC systems are available in [9].

20

These organizations publish more official documents containing data in question (annual reports, statistics summaries

etc.); however, this data is sometimes different in different documents.



PROCEEDINGS 164/208

3.1 Improvement of technical measures

Functions performed by existing LC systems are realized with very high safety integrity level (SIL4 according to [1]); however, failures of LC systems contribute to accident rate at LCs operated by the ŽSR less than 0,1%. Safety could be increased by implementation of new functions or modifications that primarily do not increase technical safety but support organizational measures and their respecting by road traffic participants, eventually help to consolidate information (extent, content, form) provided to them. Generally, safety at LCs could be increased by:

• Rigorous use of barriers ; statistics indicates that accident rate at LCs with barriers is lower than at LCs without barriers. Using barriers has also disadvantages – higher operational cost (due to frequent forced damaging of barriers) and higher time of LC closure [9]; it is necessary to consider if arguments against installation of barriers are strong enough when comparing safety of LCs with and without barriers;

• Minimum time of LC being closed , allowing the slowest and longest road vehicle safely to pass the

level crossing; if there are LCs at railway lines where rail vehicles move with different speeds the “classical control” (based on the fixed activation point) requires using speed of the fastest train in warning time calculations and particular LC configuration (e.g. crossing angle of rail and road communication, angle between barrier arm and road axis, width of road communication, width of a lane, distance of end track axes, width and length of the LC dangerous zone, etc.); this results in situation when the LC is closed uselessly long for slow trains (according to our calculations for the speed range from 60 to 100 km⋅h-1 the time LC being closed is in the case of LCs using a specialized railway signal sometimes extended as for 70%) which has a negative effect on psychics of drivers. Using approaching time predictors could solve this problem – this idea is not new but technologies available and realizable at present are more suitable than those known in the past. The basic task of approaching time predictor is to ensure the same (constant) approaching time for a certain range of rail vehicle speeds [11];

• Informing of a locomotive driver on state of the LC and minimizing probability of train coming

to unsecured LC ; thanks to information about LC failure the locomotive driver can slow down and stop in front of the obstacle. Information about the LC state may be transmitted to the main railway signal, specialized railway signal (informing about LC state only) or directly to the locomotive crew (really realizable in connection with building ETCS and GSM-R). In the case of using a specialized railway signal it is necessary to inform about functionality of the LC (not about warning state), which requires a proper design of the LC system. If the signal informs on the warning state then it must be installed inside the approach section at the braking distance and required visibility of the boundary of approach section must be ensured – that may result in extension of LC closed time. Validity of this conclusion is not general and holds good for example for LCs with full barriers since length of approach section is the same for all functions of the specialized railway signal and independent on the crossing angle between road and railway.

• Monitoring of dangerous area of the LC , as an example we could mention the use of a camera

system identifying an obstacle at the LC; however, this additional system can be meaningful only if information about obstacle is available to the locomotive driver and he/she has a chance to start intensive braking and stop in front of the LC. Such an approach requires dependable monitoring information because false message on existence of the obstacle also causes activation of emergency braking system with possible injures of train passengers [2]; implementation of such a system can only prevent collision with a car that has been broken and is standing for a certain time in the dangerous zone, or a car driven by a suicide driver, who is intentionally waiting for train coming. System seems to be ineffective when monitoring behaviour of “zig-zag” drivers who violate rules but spend very low time when crossing the LC. Automatic penalization realized on the camera-based proof can be a good enforcement measure, however in some countries may be less or more in conflict with human rights legislation.



PROCEEDINGS 165/208

• Unification of information and the way of its signa lling to road traffic participants ; the following problems exist:

o Non-uniform way of informing about failure state of the LC systems; this may be demonstrated at

the example of the ŽSR: - If a locomotive driver is informed about failure of the LC system, the warning state at the LC is not activated and from the view of road participants the LC looks open and safe; - If a locomotive driver is not informed about the failure, the warning state at the LC must be activated and the LC looks closed; from the safety point of view it would be desired to keep LC closed till the moment when a locomotive driver of the approaching rail vehicle becomes informed and such information is given in such a distance that he/she can reduce actual speed to be able to stop in front of the obstacle. However, long time of LC being closed evoke negative feelings of road traffic participants resulting in passing the LC even under the warning state;

o Non-uniform signalling of the warning state in EU countries; different kinds of LC signals are used

in EU countries; speciality of the ZSR is usage of so called “active signalling”, represented by white flashing light (implemented as a part of the LC signal); this signalling informs road traffic participants about the fact that in the LC area there is no rail vehicle that could endanger them and allows them to pass the LC with almost doubly higher speed. Since “active signalling” is not implemented at all LCs (61% out of all 976 active automatic LCs were equipped with this kind of signalling in 2006), from the safety point of view its application has more negative than positive effect. Another important aspect is that white flashing light at the LCs may be strange for foreign car drivers who are not familiar with national specificities and who may misinterpret its meaning.

3.2 Improvement of organizational measures

According to the Slovak law [13] a road vehicle driver who is approaching the LC is obliged to behave extremely carefully and make sure of possibility to pass safely the LC. This statement may sound strange to those used to pass LCs in western European countries but seems normal and acceptable in many countries of Eastern Europe. Particularly, in Slovakia in a distance 30 m to the LC a car driver is allowed to go with the maximum speed 30 km⋅h-1. Neither existence of active signalling releases him/her from liability to pass the LC safely. The law [13] says that if white light is flashing (i.e. active signalling is installed and being on), a road vehicle driver is obliged to go maximally 50 km⋅h-1 in a distance 50 m to the LC. Due to this diction of law, making organization measures more rigorous and stricter is practically impossible. Safety could only be increased by more rigorous monitoring of how traffic regulations are respected by road traffic participants. Different technologies are available to support this process (for both monitoring and enforcement functions) as mentioned above. If the LC is not equipped with the LC system, then safety of road traffic participants passing the LC is based on organizational measures only. One way or the other, at present all LCs at the ŽSR must be equipped with St. Andrew’s Crosses and a group of traffic signs informing about approaching the site of LC. One of recommendations considered and discussed within the SELCAT project was recommendation to unify the role St. Andrew’s crosses and ensure the same meaning all over the Europe. This measure could be implemented at minimum cost (unlike very costly and hardly realizable unification of other LC signalling). The main idea is that St. Andrew’s Crosses should be installed at those LCs only where railway operator does not guarantees safety and a road traffic participant is obliged to mind his/her own safety when passing it.

4. Conclusions

Safety at Slovak LCs could further be increased by the following measures: • Installing of active automatic LCs systems with barriers; • Increasing availability of the LC systems – this measure will cause extension of organizational

measures to the railway staff, too; • Stricter monitoring of how organization measures are respected, especially by road traffic participants

and enforcement (e.g. equipment based on camera recording and fining those participants who violate traffic regulations);



PROCEEDINGS 166/208

At the same time we can conclude that:

• Increasing of safety at LCs must be solved mainly on the side of road traffic participants; • Further increasing of technical safety of LC systems is not necessary; • Use of “active signalling“ at the ŽSR does not contribute to higher safety and may be a cause of

potential misinterpretation by foreign road traffic participants; • Present version of the law 315/96 Z.z. is more preferable for railway operator than to road traffic

participants. Analysis of statistic data about accident rates at LCs may:

• Provide data necessary for risk analysis [4] and define requirements for technical safety of newly developed LC systems;

• Help to optimize range of technical and organizational measures to ensure safety at LCs. Nowadays corridor lines are in the process of significant reconstruction all over the Europe. Within efforts for interoperability the European Parliament adopted the Directive [6] to guarantee minimum safety level of railway transport in member state of the EU. This aim could be achieved only provided that a single methodology is elaborated to:

• Record and appraise statistic data about accident events; • Analyze risks and define safety requirements; • Assess safety.

Acknowledgements

The paper was prepared with support of the 6FP No. TCA-CT-2006-031487 project “SELCAT - Safer European Level Crossing Appraisal and Technology” and partial support of the 6RP/SELCAT (MVTS grant).

References

[1] EN 50 129: Railway Applications: Safety related electronic systems (2003). [2] Grimm, M.; Pelz, M.; Meyer zu Horste, M. “Safety Relevant Applications at Level Crossings by Means of

Imaging Methods“, Proc. of the 15th int. symposium EURNEX-ZEL 2007, 2nd part, pp. 39-46 (2007). [3] Internet source: http://europa.eu.int/comm/transport/era/doc/wp2005.pdf [4] Rástočný, K.; “Risk Analyses for Railway Signalling System. AEEE Journal, No. 3-4, Vol. 2 pp. 24-29

(2003). [5] Rástočný, K.; Janota, A.; Zahradník, J. ”Safety at Level Crossings of ZSR”, Proc. of the 15th int.

symposium EURNEX-ZEL 2007, 2nd part, pp. 77-84 (2007). [6] Safety Directive 2004/49/EC of the European Parliament and of the council of 29 April 2004 Available at: http://eur-lex.europa.eu/LexUnServ/site/en/oj/2004/1_220/1_22020040621en00160039.pdf [7] SELCAT – Annex I: Description of Work. Proposal No. 031 487 (28th April 2006). [8] Strategic Rail Research Agenda – Technical Annex, ERRAC (2002). Available at: http://www.errac.org/docs/ERRAC_SRRA_Tech_Annex.pdf [9] STN P 34 2651: Železničné priecestné zariadenia (1999). [10] White Paper: European Transport Policy for 2010: time to decide (2001). Available at: http://ec.europa.eu/transport/white_paper/documents/doc/lb_texte_complet_en.pdf [11] Zahradník, J.; Rástočný, K.; “Aplikácie zabezpečovacích systémov”, EDIS – publisher ZU (2006). [12] Zahradník, J.; Rástočný, K. “Sicherheit des Verkehrs an Bahnübergängen der ZSR“, Signal und Draht, No.

6, pp. 22-25 (1998). [13] Zakon c. 315/96 Z.z. o premávke na pozemných komunikáciách a s ním súvisiace vyhlášky a predpisy [14] Z1: Pravidlá železničnej prevádzky (2005).



PROCEEDINGS 167/208


Bentleigh, Victoria – Australia - a Railway Pedestr ian Crossing Case Study

Author(s): Terry Spicer

Job Title: Manager Railway Crossing Safety

Company: Department of Transport - Victoria. Public Transport Division.

Safety & Asset Management Branch

Country: Australia Resume of Speaker Terry Spicer has been involved in train operations management in the public and private sectors of the Australian rail industry for over 47 years, including 39 years with the Victorian Government and 8 years in the iron ore industry in the Pilbara Region of North West Australia. Terry advises the Minister for Public Transport on all matters pertaining to railway crossing safety in Victoria and represents the Department of Transport at Coronial Inquests related to railway crossing fatalities. Terry provides the Secretariat role to the Victorian Railway Crossing Safety Steering Committee and represents the Director on the Victorian Railway Crossing Project Delivery Group, the Victorian Railway Crossing Technical Group and the Victorian Railway Crossing Safety Awareness Committee, as well as representing Victoria on the Australian Level Crossing Assessment Model (ALCAM) National Committee. Abstract



PROCEEDINGS 168/208

This paper summarises a Case Study of enhancements put in place at the Bentleigh, Victoria, Australia, railway pedestrian level crossing for disabled user groups as well as trials of additional pedestrian crossing safety features. The main drivers for these enhancements were to provide a safer railway pedestrian crossing environment for all railway pedestrian crossing users, and to comply with legislative requirements of the Commonwealth Disability Discrimination Act (DDA) 1992 and the Disability Standards for Accessible Public Transport Amendment 200421. Some of the issues that were considered during the development phase were:

• The appreciation of the risk to the various user groups. • The development of standards and user groups acceptance. • Challenges to construction.

The appreciation of the risks to the user groups The risks for users at pedestrian level crossings were discussed and documented at numerous nationwide workshops involving disabled users and transport groups. These workshops were the forums for open discussions that provide the foundation for the development of options to address the risk for disabled users at pedestrian crossings. The development of standards and user groups accept ance The development of the new Victorian Rail Operators Group (VRIOG) and Australian Standards were initiated by the construction of a full size replica of the proposed swing gate pedestrian crossing at the Yooralla Independent Living Centre, Brooklyn Victoria . Various disabled groups and representatives of various rail and road transport bodies, participated in the evaluation and assessment of the proposed swing gate pedestrian crossing configuration which has been adopted in both standards. Challenges to construction Pedestrian level crossing safety improvements are becoming more complex as technology becomes more sophisticated and user needs increase. Challenges to meet DDA standards may include:

• An appreciation of gradient approaches to ensure that the finished gradients comply with the 1:14 ramp DDA standards.

• Special width (2.44m) and non-slip, flat, even, surface materials of walkways. • The provision of TGSI (Tactile Ground Surface Indicators) for delineation and hazard indication for the

vision impaired. • Increasing the standard size of the maze to accommodate mobile disability devices like Gophers and

Scooters etc. • Enhanced visual and audio warning devices for the sight and hearing impaired. • Mechanical (Victoria)/Electromagnetic (New South Wales) emergency escape gate latch. • Mitigation of SPAD (Signal Passed at Danger) issues at level crossing. • Improved pedestrian lighting on a risk assessment basis.

1. Introduction Bentleigh Railway Station (16.6 km south east of Melbourne) on the Frankston electrified metropolitan rail line, is adjacent to Centre Road, which crosses the railway line at-grade immediately south of the Station. In the mid-1990’s the railway pedestrian subways ( constructed as part of the provision of the third bi-directional rail line in 1978) on both the Melbourne (north) side and Frankston (south) side of the Centre Road Bentleigh railway crossing were filled-in at community request, due to having a record of physical (including sexual) assaults, drug offences, vandalism, flooding and poor lighting. At that time pedestrian access to the station island platform on the north side was available either via the subway or an at-grade pedestrian crossing. The south side pedestrian access also included both a subway, (which only provided grade separated crossing of the rail line), as well as an at-grade pedestrian crossing. Subsequently, in the period between 23 March 1998 and 19 November 2004 there were three separate pedestrian fatalities at this location. The first one of which involved a (18 year old male) pedestrian, who on 23

21

Spicer Terry – Fatal Accidents and Disability Access at Pedestrian Railway Crossings in Victoria, Australia – 2004.



PROCEEDINGS 169/208

March 1998, is suspected to have either used the road crossing, used the emergency escape gate, or climbed over the pedestrian barrier, whilst wearing a ‘walkman’ style transistor radio. The second one of which was a (25 year old female) student who on 6 December 2000, forced open two closed pedestrian gates. The last one of which was due to a (15 year old female) student using the emergency escape gate to access the track. In all three fatalities it was ‘illegal access’ and the ‘second train syndrome’ which were the common threads and which contributed to the fatal accidents. In between the second and third Bentleigh fatalities we also had two fatal wheelchair accidents at Nunawading in October 2001 and at Clayton in December 2001, which were featured in the paper22 presented to the 8th International Level Crossing Symposium in Sheffield in the UK in 2004. Immediately following the third fatal Bentleigh pedestrian accident there was mounting public, media, and political pressure to reinstate the closed pedestrian subways. On 30 November 2004 the Director of Public Transport advised the Minister for Public Transport that a review of the pedestrian crossing at Bentleigh Station would be conducted to assess:

(a) the feasibility/practicality and cost of re-opening the pedestrian underpass and ramps on the north side of the Station, by addressing the problems which led to its closure and requirements which could make it a safer alternative;

(b) whether additional signage and/or warning devices are required at the existing pedestrian crossing; and

(c) whether emergency exit gates can be designed to remove the ability for them to be opened (by pedestrians) from the ‘wrong side’.

The review was required to make recommendations in respect of (b) and (c) whether or not re-opening of the pedestrian subway was found to be the better solution. On completion of its preliminary investigations the review team engaged ARUP Pty Ltd (Arup), Consulting Engineers to undertake surveys of public behaviour at selected pedestrian rail crossings, including those at Bentleigh and to evaluate options for providing grade separated access across the rail tracks at Bentleigh. The final review of Bentleigh did not recommend reinstatement of the pedestrian subways, as the practical disadvantages of each would require that both pedestrian crossings be retained in the event that any one of the options was implemented. Surveys indicate that where there is a choice between at-grade and underpass access a significant majority of pedestrians will choose the at-grade option which is the shortest, quickest and most convenient route across the railway line. The report also concluded that the option to lower the railway line and to bridge Centre Road over it is not achievable in the foreseeable future given the cost and extent of works involved. Given that both pedestrian crossings need to be retained at-grade, it is important that all practical steps are taken to improve the level of safety for users. To this end it was recommended that: - treatments specified in the draft update to Australian Standards AS1742.7-1993 (issued in February 2007)

and the VRIOGS 003.2-2004 (disability access layouts, track) being trialled on at-grade pedestrian crossings at Carnegie, Kensington, and Thornbury, to meet requirements for people with disabilities and which deliver improvements for all users, be applied to both pedestrian crossings at Bentleigh by 31 December 2005.

- in addition fencing and gates at both crossings at Bentleigh be upgraded to the standards specified in

updated draft Standards AS1742.7-1993 and VRIOGS 003.2-2004 (draft) by 31 December 2005. - both crossings at Bentleigh be used for the field trialling of the Surite pedestrian gate latch for emergency

exit gates which is being factory-tested at present; subject to satisfactory factory-testing field trials will commence by 31 December 2005.

- signage consistent with draft update to Australian Standard AS1742.7-1993 for exit gates on emergency

enclosures be implemented on both crossings at Bentleigh by 31 December 2005.

22




PROCEEDINGS 170/208

- Red Man and Second Train Crossing electronic signage activated by approaching trains be installed on the station side crossing at locations between the up and centre rail tracks and east of the down track as soon as type approval is achieved and that Bentleigh be used for trials commencing 31 December 2005; and

- current public transport education and information programs conducted by the Metlink Transport Education

Service and Connex Melbourne Limited, emphasising rail pedestrian crossing safety, be promoted and reinforced in schools which have students using Bentleigh Station.

The recommendations contained in the review were adopted by the Government and the Bentleigh pedestrian crossings were subsequently upgraded as part of a $1.2M package. The package included an upgrade of the crossing layout, surfaces, and design, in order to comply with the review recommendations listed above. Due to the size limitations imposed upon papers to Level Crossing 2008, it is not possible to include the full Bentleigh Case Study, which contains details of the Bentleigh Railway Crossing Review - 2005, photographs and diagrams in all of the documents, including the Sinclair Knight Merz, Centre Road Bentleigh Concept Design – Criteria Assessment - 2005, the VRIOGS 003.2-2004 (draft) standards, as well as the ARUP Consultant Pedestrian Behaviour ‘Before & After’ studies 2005, 2007 and 2008, and implementation of the second to fifth dot point safety treatments (immediately above) related to the red man/another train warning sign trials and the emergency escape – gate latch trials. The remainder of this paper summarises the contents of the Case Study. 2. Bentleigh, Victoria – Australia; A Railway Pedes trian Crossing Case Study The Bentleigh Case Study consists of combining all of the following documents which are summarised in Sections 4 to 11 of this paper and included in the references in Section 12. A full CD version of the “Bentleigh, Victoria – Australia; A Railway Pedestrian Crossing Case Study,” can be obtained from the author at the above email address. 3. Railway Pedestrian Crossings Centre Road Bentlei gh, Review – July 2005 As a result of the third fatality on 19 November 2004 the following was announced: 4.1 Review of Railway Pedestrian Crossings – Bentleigh - July 2005 - Terms of Reference. On 30 November 2004 the Director of Public Transport advised the Minister for Transport that a review of the pedestrian crossings at Bentleigh Station would be conducted to assess:

(a) the feasibility/practicality and cost of re-opening the pedestrian underpass and ramps by addressing the problems which led to its closure and requirements which could make it a safer alternative;

(b) whether additional signage and/or warning devices are required at the existing pedestrian crossing; and (c) whether emergency exit gates can be designed to remove the ability for them to be opened (by

pedestrians) from the ‘wrong side’.

The review was to make recommendations in respect of (b) and (c) whether or not re-opening of the pedestrian subway was found to be the better solution. It was agreed that the review would be led by the Department of Infrastructure (DOI) and that the Director, Public Transport Safety and the Glen Eira City Council would be consulted. On 3 March 2005 Rob Hudson MP, Member for Bentleigh, referred concerns raised by a constituent about the safety of the pedestrian crossing on the south side of Centre Road and requested that it be examined as part of the review of the crossing on the station side of Centre Road. It was agreed that this should occur and that a report covering both crossings would be completed.



PROCEEDINGS 171/208

4.2 Pedestrians Struck by Trains Interrogation of the databases from 1987 to 2005 indicated that there have been three fatal pedestrian accidents at Bentleigh: 7. 40 am - 23 March 1998: Bentleigh (station side crossing). The State Coroner found that the deceased (entered the crossing from the eastern side), ignored signals and was struck by a Melbourne-bound train approaching from the south (on the centre track). The investigating Police Officer stated that “the deceased either went through the emergency gate displaying a NO ENTRY sign, opening this unlockable gate and proceeded through the pedestrian gate or went on to Centre Road and went around the pedestrian barrier, or the deceased may have climbed over the pedestrian gate that closed automatically in line with the (road) boom-gates”. 8. 43 am - 6 December 2000 : Bentleigh (station side crossing). The State Coroner found that the deceased entered the crossing from the western side, disobeyed warning signs, forced open the NO ENTRY or closed automatic gate and walked into the path of an oncoming (Melbourne-bound) train (on the centre track). 8. 45 am - 19 November 2004 : Bentleigh (station side crossing) The Coroner found that the deceased died from severe head injuries sustained when she unintentionally walked in front of and was hit by a north-bound express train. The Coroner also noted that the deceased opened the unlocked pedestrian emergency exit gate which had a sign on it saying "No Entry". Despite the operating bells and lights and the warning train whistle, she entered the railway line area. She seemed to remain unaware that a train was approaching as she began to walk across Line 2. There were several recommendations related to the findings. Nearing completion of its preliminary investigations the review team engaged Arup Pty Ltd (Arup), Consulting Engineers, to undertake surveys of public behaviour at selected pedestrian rail crossings including those at Bentleigh and to evaluate options including costs for providing grade separated access across the rail tracks at Bentleigh. The report details the investigations undertaken, addresses each term of reference and makes a number of recommendations. These recommendations were adopted by the Government and the Victorian Rail Track Corporation (VicTrack) were appointed Project Managers to implement the recommendations. 4. Arup - Platform Access At Bentleigh, Carnegie an d McKinnon Stations – June 2005 Pedestrian counts at three locations were undertaken to determine the level of dangerous crossing activity currently occurring at Bentleigh, McKinnon and Carnegie Stations. McKinnon Station has both an at-grade crossing and an underpass while Carnegie station (on the adjoining Dandenong rail line) has an underpass on one side of the road and an at-grade pedestrian crossing on the other side. Arup was commissioned to assess the level of pedestrian crossing activity currently occurring at Bentleigh, McKinnon and Carnegie Stations and this report provides an analysis of pedestrian behaviour, including illegal crossing behaviour, at these three locations, for comparison purposes. 5. SKM - Bentleigh Centre Road Concept Design – Criteria Ass essment – 6 September 2005 VicTrack commissioned Sinclair Knight Merz to develop a scope of works and concept design for upgrading the pedestrian gates and associated infrastructure at Centre Road Bentleigh. As part of the development VicTrack sought assessment of treatments currently under trial or review for inclusion in the concept design. The treatments to be considered come from four key sources. They are;

• Draft Update to Australian Standard AS1742.7 – 1993 • Draft Standard VRIOGS 003.2 – 200423 • Trial installations at Carnegie, Kensington and Thornbury24

23




PROCEEDINGS 172/208

• Gate latch trials at the Independent Living Centre – Brooklyn.

The SKM report assesses the applicability of the treatments and guidance for the concept design. 6. SKM - Bentleigh Centre Road Signalling Concept D esign Assessment – 27 September 2005 This brief was also commissioned by VicTrack to investigate train signalling options to reduce the time the motorised pedestrian gates are closed to pedestrians, leading to shorter waiting times and less risk taking behaviour, including the practise of bypassing the motorised gates, by walking through the emergency escape gate or via the road. 7. Bentleigh - DDA Compliant Pedestrian Crossing Wo rks – VRIOGS 003.2 – 2004 (draft) The Victorian Rail Industry Operator’s Group (VRIOG) were established (post privatisation of public transport in 1999) for the purpose of establishing standards which, if implemented throughout the Victorian Rail Network, will facilitate the interoperability of infrastructure. VRIOGS 003.2 ‘Criteria for Infrastructure at Railway Level Crossings – Pedestrian Crossings’, was initiated in 2002, immediately following the two fatal wheelchair accidents in Victorian in late 2001. They incorporate the Commonwealth DDA legislative requirements outlined in the paper presented to the 8th International railway crossing symposium in 2004 and were applied at Bentleigh. Due to the topography on the south side of the Centre Road level crossing the DDA application included 1:14 gradient approach ramps for the mobility impaired. At the time of construction VRIOGS 003.2 – 2004 was the draft version applied. The final version VRIOGS 003.2-2006 is included in the case study with before and after photographs of the Bentleigh pedestrian crossings. The physical works included;

• Installation of 2440 mm wide, non-slip, moulded rubber insert pedestrian pathways. • Installation of Tactile Ground Surface Indicators to DDA compliance standards. • Installation of an additional independent automatic pedestrian gate (previously only one gate was

present in the centre refuge with holding pens on either side). • Construction of the new South West crib including a new DDA ramp and steps. • Wider escape areas and emergency gates to accommodate Gophers/Scooter mobility devices. • Upgrade of signage to comply with the latest draft standard. • Increase in fence height.

8. ARUP - Bentleigh Station Upgrade – Pedestrian Af ter Counts – March 2007 This report details the “after“ pedestrian crossing movement counts undertaken to evaluate what proportion of pedestrians cross the tracks illegally since the new signs and safety features were been installed in December 2005. A comparison with the previous (before) survey undertaken on Tuesday 21st June 2005 has also been made. 9. VicTrack – Bentleigh Interim Report – 19 Novembe r 2007 (To Be Updated June 2008) The purpose of this report is to provide an interim assessment of the trial of the various technologies installed at the Centre Road, Bentleigh pedestrian crossings. The packages include an upgrade of the pedestrian crossing layout, surfaces and design, to comply with Disability Access Standards. The site was also used to trial three new technologies designed to improve safety. These technologies are:

• Another Train Coming active warning signs, • Red Man active warning sign, and • Emergency Escape Gate Latches – to prevent illegal access.

The trials were designed to not only test the equipment and how suitable it was for type approval, but also to test how effective the new measures are at achieving their intended purpose. There are six ways in which information is being collected in order to determine this:

• Before and after studies of pedestrian numbers and illegal crossings • Feedback from Industry (train drivers, maintenance staff).

24




PROCEEDINGS 173/208

• Feedback from the general public through signs installed onsite. (Facing cover page of this paper). • Information collected through the use of CCTV cameras installed to monitor the crossing • Information collected as part of routine inspections • Thorough inspection of the equipment and internal components

The scheduled completion of the trial was originally the 31st July 2007 however, due to the well publicised braking faults with the Siemens train fleet in Melbourne it was extended till the 30th September, 2007. This was due to the operation of level crossings adjacent to railway crossings being altered, extending the amount of time that the crossings were active as a temporary safety measure. It was accepted that this may skew any attempts to determine how successful the technologies have been at achieving their intended purpose, as there is a correlation between long crossing closure times and the increased numbers of illegal crossings. It was agreed at the October 2007 meeting of the Railway Crossing Technical Group that the trial would be extended until March 2008, at which time the after study will be conducted again (see 11 below). This date coincides with university reconvening and provides a number of months of normal crossing operation for pedestrians to display natural behaviour for the study. The report provides information on the pedestrian crossings design and layout, DDA applications, an overview of each technology, performance of each, areas for improvement, and conclusions and recommendations. 10.1 Summary of Conclusions and Recommendations The technology has been effective at reducing the percentage of illegal crossings at Centre Road. This is supported by comments from the rail industry that fewer people are taking risks as well as the report’s completed by Arup on illegal pedestrian movements after the upgrade. Another Train Coming Issue: Action Phantom Aspect • Review current design face

• Risk assess phantom aspect Test Switches • Install test switch for ATC at Bentleigh Size of sign • Review size of sign ATC Operation • Investigate ATC signs not activating for a train that trips past a signal LED Module Failures • Future modules to be flow soldered, Mounting and Wiring • Design standard wiring layout and mounting brackets

• Retrofit Bentleigh where appropriate Recommendation: Implement actions for each issue and modify Bentleigh as required and retrial. Red Standing Man Issue: Action IP65 Rating • Determine required IP rating through signalling guidelines, modify signs as

required LED Module and Power Supply

• Investigate combining LED module and power supply on one PCB

Recommendation: Implement actions for each issue and modify Bentleigh as required and retrial. Emergency Escape Latch Issue: Action Electromagnetic latch vs. Mechanical latch

• Create a report comparing both technologies including full lifecycle cost. Submit to Railway Crossing Technical Group for approval.

Engineering concerns • Resolve pending outcome of comparison Wrong side entry • Reposition CCTV Camera over North Eastern Escape area to ascertain how

pedestrians are opening the gate from the wrong side Recommendation:



PROCEEDINGS 174/208

Ensure the latches onsite continue to operate as designed through regular inspections and implement the above actions. Modify Bentleigh as required and retrial. Spring Closure and Hinges Issue: Action Rate of Closure • Evaluate speed of closure from spring Recommendation: If mechanical latch used, develop a closure device that will control the rate of closure (This will not be a concern with the electromagnetic latch due to the weight of the gate). Other Areas of Improvement Issue: Action Height of Fencing • Approve the use of 1.5metre fencing height for future pedestrian crossings where

latches are to be provided. Emergency Escape Signage

• Review current signs and ascertain why confusion exists.

Recommendation: Modify Pedestrian Crossing Standard to reflect changes to fencing height and evaluate emergency escape signs. 10. Arup – Bentleigh Rail Crossing Pedestrian Safet y Features Trial Site – Survey of Pedestrian

Crossing Movements April 2008. This report details the latest “after“ pedestrian crossing movement counts and provides a comparison with previous surveys undertaken in June 2005 and February 2007. The new survey of illegal pedestrian movements has been requested following a review of ringing times i.e. boom gates down at the station, in respect to trains travelling between Bentleigh and Patterson Stations. The 27 illegal movements recorded over the six hour period of the latest survey was a reduction on the 34 crossings noted in 2007 and the 30 illegal crossings in 2005. Of the 27 crossings recorded this time, 19 (70%) occurred in the AM and eight (30%) in the PM; with 21 occurring on the north side and six on the south side. The volume of pedestrians using the crossing has increased markedly over the three surveys and the percentage of illegal crossings has continued to drop. The percentage of illegal crossings in the AM period was 1.21%, similar to 2007. The number has reduced due to six more pedestrians being counted with the same number of illegal crossings. In the PM period the level of 0.29% has significantly reduced compared to the 0.64% recorded in 2007 and the 0.78% recorded in 2005. For the overall six hour period, illegal crossings constituted 0.62% of pedestrians (27 of 4,247 crossings), down from the 0.87% average recorded in 2007 (34 of 3,912 crossings) and the 0.98% average recorded in 2005 (30 of 3,060 crossings). The illegal movements noted occurred primarily via the bypass gates. Observations of the occurrences showed that the majority of cases involved single instances of pedestrians. Of interest in this survey was the increase in illegal movements on the south side with six overall – two in the AM and four in the PM; all using the south side bypass gates. We believe that the illegal movements via the bypass gates are people (particularly school children) who are subverting the emergency escape gate egress release buttons by various means, which have been positioned for both mobility and non-mobility users to legally exit the crossing, in the event of becoming caught on the crossings when the active controls commence operating. As a consequence, at the time of writing this paper we are currently re-positioning the CCTV camera’s to more closely monitor pedestrian behaviour, in order to develop counter illegal access strategies.



PROCEEDINGS 175/208

11. References [1] Arup, “Platform Access at Bentleigh, Carnegie and McKinnon Stations - Survey and Assessment”, Internal

DOI Document, (June 2005). [2] Arup, “Bentleigh Station Pedestrian Upgrade - Pedestrian After Counts”, Internal VicTrack Document,

(March 2007). [3] Arup, “Bentleigh Rail Crossing Pedestrian Safety Features Trial Site – Survey of Pedestrian Crossing

Movements – April 2008”, Internal VicTrack Document, (April 2008). [4] Hogan, C. VicTrack, Spicer T. DOI, “Bentleigh, Victoria – Australia; A Railway Pedestrian Crossing Case

Study”, Internal VicTrack/DOI Document, (April 2008). [5] Lyons, G. (VicTrack); Sargant, T, (DOI/PTD); Trevaskis, R. (Connex ML); “Railway Pedestrian Crossings

Centre Road Bentleigh”, Internal DOI Document, (July 2005). [6] Sinclair Knight Merz, “Bentleigh Centre Road Concept Design – Criteria Assessment”, Internal VicTrack

Document, (September 2005). [7] Sinclair Knight Merz, “Bentleigh Centre Road Signalling Concept Design Assessment”, Internal VicTrack

Document, (September 2005). [8] Spicer, T.; “Fatal Accidents and Disability Access at Pedestrian Railway Crossings in Victoria, Australia”,

Eighth International Level Crossing Symposium, (April 2004). [9] VRIOGS 003.2 – 2004 (draft) “Criteria for Infrastructure at Railway Level Crossings – Pedestrian

Crossings”, Internal DOI (draft) Document, (2004).



PROCEEDINGS 176/208

Session 9 - Engineering and Operation Chairman: M.C. Murali, Great Southern Railways, Ind ia

Customised techniques and operational rules to imp rove level crossings by means of imaging methods

Author(s): Markus Peltz

Job Title: Researcher

Company: German Aerospace Centre, Institute of Transportation Systems

Country: Germany

Resume of Speaker

Dipl.-Ing. Markus Pelz studied Transport and Transportation Systems with the main focus on railway safety at the Technical University of Dresden (Germany) at the Chair of Railway Signalling and Transport Safety Technology. After having finished his diploma thesis in 2005, he started to work at the Institute of Transportation Systems (ITS) at the German Aerospace Centre (DLR) in Braunschweig. There he is the expert in level crossing safety systems. Markus gained his first professional experiences of level crossing technology from his industrial placement at Scheidt & Bachmann. The main focus of his job is to find out new methods to improve level crossing safety systems.

Abstract A significant part of the existing level crossings were built for cart-tracks or streets with few traffic in the near of crossroads of main streets. The characteristic of most of these level crossings is laying in the orthogonal adjustment to the main street. The operationally most efficient alternative is the use of an on-call barrier. But this technology can only be realized with cost-intensive common of the shelf components. To show the possibilities to automate an on-call barrier system to reduce costs without neglecting safety, the Institute of Transportation Systems (ITS) of the German Aerospace Center (DLR) will build up a demonstration system to analyze the alternatives in realizing technical protection of level crossings in economically efficient ways without reducing the safety. In addition to the on-call barrier function different other functions, like danger zone supervision, will be developed. Therefore the ITS will build up customized techniques, and operational rules to improve level crossings by means of imaging methods will be analyzed.

Introduction

The Institute of Transportation Systems (ITS) of the German Aerospace Center (DLR) in Braunschweig investigates the situation of secondary lines in present and future in Germany. This project is funded by the Ministry of Economics in Lower Saxony. In particular, technical and operational solutions, which result in cost reducing improvements for the operating company, will be analyzed. One approach to increase the economical situation without neglecting safety aspects is found in the adoption of imaging methods for safety relevant applications in railways.



PROCEEDINGS 177/208

In the past, several approaches were carried out to build up automated railway operations using imaging methods. Most approaches were using video based camera technologies within optical systems. However, none of these approaches were implemented in practice. On the contrary, imaging methods for applications to assist the operations of the German railways are in use since a longer period of time. As an example may be named that the Hamburg commuter railway system is using video technology for the dispatching of the trains in stations by the train driver. Further applications for imaging methods for example are the track surveying and the level crossing monitoring. The video based level crossing monitoring is only used as a technical aided system for the signal man to observe a distant level crossing. The protection of the level crossing against unauthorized crossing is still not implemented using imaging methods. In Germany, a significant part of the existing level crossings were built for cart-tracks or streets with few traffic in the near of crossroads of main streets. The characteristic of most of these level crossings is laying in the orthogonal adjustment to the main street. The operationally most efficient alternative is the use of an on-call barrier. But this technology can only be realized with cost-intensive common of the shelf components (see fig. 1). A high number of not technically protected level crossings exist today worldwide without any obstacle detection in different structural and operational constellations. Therefore the Institute of Transportation Systems of the DLR investigates the options for technically realized vacancy detection for the danger zone of level crossings by a cost efficient way.

Fig.1. typical constellation of an on-call barrier at cart-tracks

With this vacancy detection it will be shown, how it is possible to close gaps in today’s safety concepts of the level crossing system by an efficient way. The goal is to improve the safety at level crossings. Because of this case, a start will be made by analyzing accidents at level crossing to find out the root causes. Within the knowledge of what the root causes are, we are able to eliminate these things by interposing new technologies, e.g. by means of imaging methods in combination with operating rules. To reach this goal, the ITS started by a progressive step-by-step development of functions at the level crossing system. These functions will be realized by optical sensors and image processing (optical systems) in combination with operational rules:

• „Unclosing barrier“: automatic opening of the barriers at full-barrier level crossing with on-call functionality. The barriers will open only when traffic wants to cross the railroad.

• „Observe closing barriers“: to avoid closing of barriers when obstacles like trucks are beneath the barriers at the same time (see fig. 2)

• „Real-time photo telegraphy to the train driver“: transmission of the situation at the level crossing to the train drivers desk. The train driver is able to see what is happening at the time when he is going to reach the danger zone.



PROCEEDINGS 178/208

• „Real-time photo telegraphy to the operator“: transmission of the situation at the level crossing to the operator. The operator sees the danger zone and its surroundings. Now he is able to, e.g. warn the train driver.

• „Obstacle detection“: to inform, to warn, to brake the train when any obstacle is between the half barriers. • „Danger zone supervision“: to make sure that no obstacle is inside the danger zone between the full-

barriers. This is a safety critical function and should provide a safety integrity level (SIL) of 2 or even higher.

Fig.2. collision of truck with a barrier

To show the possibilities of a technology based on optical sensors, the ITS of the DLR will build up a demonstration system to analyze the alternatives in realizing technical protection of level crossings in economical efficient ways without reducing the safety.

Motivation

All over Europe there is a multiplicity of technically secured level crossings (see fig. 3). Though the chance of an accident at a level crossing (LX) according to other accident hotspots is very low, there are numerous incidents at LX with very high measures of damages [1]. Furthermore, it can be said that as a result of different appearances of the LX road securing system the car driver is confronted with a system at LX with very high complexity where it is not relevant whether the LX is equipped with semi-barrier or only secured by flash lights, because the car driver will ride over a secured LX anyway without any attention to the trackside of the LX. In this contribution it will be shown how a LX can be designed with more performance for the LX securing system, when the roadside is included in the whole system design.

Fig.3. Level crossing with half barrier and light signal



PROCEEDINGS 179/208

Performance vs. safety

In many countries, LX on less important roads and railway lines are often open or uncontrolled, sometimes they are equipped with warning lights or bells to warn the car driver of approaching trains. LX without barriers represent a safety issue. Many accidents have occurred due to failure to notice or obey the warning. In the German Allgemeines Eisenbahn Gesetz (AEG) it is said that "Railways in Germany are obliged to build their vehicles and infrastructures in a safe way and to keep them in a safe state." [2] To reach this requirement, it is common practice to learn out of dangerous situations, incidents and accidents to identify weak spots of a system and eliminate them. This contribution shows how the system safety of a complex structure like that of a LX can be increased by the use of non-common methods. This could in future lead to the development of a new LX securing system. In rail traffic it is necessary to take special technical and operational measures for realising reliable and safe rail operations because of the longer braking distances in comparison to rail traffic and the missing possibility of a train to avoid. Such measures are resulting in higher operational costs although the railway operators are under increasing cost pressures. In Europe a lot of LX systems are secured for the road traffic only by a LX warning sign (see fig. 4). This is not really performed to the operation and to the safety in railways. Additionally there is no system, which allowes the train driver to react in urgent cases of a dangerous situation.

Fig.4. Level crossing without barrier and light signal

Expensive technology vs. Economic interests

Because of system inherent features of the railway, trackside equipment is exposed to high stress resulting out of climate, vibration and electromagnetic radiation. Thereby, maintenance works with high financial and personnel efforts are resulting. The initial costs of a system that resists these circumstances are very high at the moment, so that the investor avoids such a capital expenditure for LX systems. One step for lowering the costs is the reduction of cabling. Furthermore it is to check, whether highly available low-cost technology can lead to a reduction of existing safety components, like expensive vacancy proving system for the danger zone of a LX, or not. The relocation of technology from the track to the on-board side can be seen as one possible way to get a cost minimisation, because special maintenance services do not need to take place, due to periodic vehicle maintenance and to reach a adaptively of the equipment to the volume of traffic.

State of the art

Today, some technical systems based on video technology are involved in the operational process of the German railway system, e.g. for the operator to watch the danger zone (see fig. 5 a).



PROCEEDINGS 180/208

In Hamburg, Germany, a video based system is in use by the Hamburg commuter railway system. The train driver obtains part of the information required for the train dispatching procedure by means of wireless video transmission. Information about what is happening on the platform is transmitted from the cameras installed on the platform to the monitors in the driver cab (see fig. 5 b) [3].

Fig.5. a. Video system at level crossing with full barrier and

b. Driver-dispatch system of the Hamburg commuter railway system

Innovative approach

In general, special signals are given to the train driver by the interlocking if a LX flash light system is faulty. This linking between interlocking and train is highly expensive and forms the main part of the total costs of a LX, though used only a few times in a year. This is why infrastructural technology should be turned down, especially on low frequented lines. Therefore, in the Switzerland an innovative system is under test, which secures a LX only by flash light in combination with a dynamic road sign instead of expensive barriers (see fig. 6). The road sign will flash yellow in case of a fault in the flash lights of the LX and the crossing of the tracks is on own risk because the train driver does not know if the LX flash light is in operation or if it is faulty.

Fig.6. An innovative LX system with light signal at Emmental [4]

Range for methods of resolutions

The answer of the above mentioned problem could be seen in low cost technologies like imaging methods. For realising a LX securing system, modern imaging methods by using optical sensors (e.g. cameras in visible an infrared range) are investigated. These optical sensors will be installed in such a way that an automatically detection of the road traffic (e.g. pedestrians, bicycles, cars, etc.) and by this an activation of the LX control can be realised. This method of resolution and a lot of other ones can be situated in the range that figure 7 shows.



PROCEEDINGS 181/208

Fig.7. range for methods of resolution for customize technique

The following functions shall be achieved by such a method of resolution to implement efficient and cost optimised rail operations, especially on secondary lines:

• safe technology with higher efficiency • extension of existing safety concepts and technology to reach better performance • safety optimisation • minimisation of harms • cabling reduction • safe low cost vacancy proving of LX danger zone

Several applications can be found in the field of railways and especially in the area of level crossings, e.g. the vacancy proving of the danger zone or the transmission of live video streams from the LX to the rail vehicle. Regarding to this contribution, only the methods of resolution for performing a vacancy proving of the danger zone of a LX is shown in detail.

The Janus Head algorithm

The optical sensors (e.g. cameras) are mounted at the LX warning sign (see fig. 8). They reduce the costs by disclaiming earth moving. The construction is called Janus Head, which means that a optical system, consisting of two optical sensors, is able to view in two different directions. One optical system means two optical sensors (camera in visible and infrared range) and computers for image processing. A Janus Head system itself consists of two optical systems (see fig. 8).



PROCEEDINGS 182/208

Fig.8. Example of operation of the Janus Head algorithm To perform a vacancy proving of the LX danger zone, a safe detection of every obstacle in the danger zone is required. The Janus Head system uses the fact that an obstacle like a vehicle first has to approach the LX from the road side before it can enter the danger zone. The approaching traffic can be detected by the used method. In a next step an algorithm can perform a vacancy proving for the danger zone by generating expectation values, which were communicated between the sensor systems and the system algorithms respectively. The example which is discussed in this contribution can include the following action sequence (or see also fig. 9):

• Camera 1 detects a vehicle and safes a picture (image 1) of the front side of the vehicle (see fig. 8) and sends a message to the LX safety system that a vehicle is approaching.

• When image 1 is send, an expectation value will be send to camera 3. • Camera 3 makes a picture at t0 of the free danger zone (image 2) and safes it. • When the barriers are open camera 2 will be activated and has to expect a vehicle. • Camera 2 detects a vehicle and safes a picture (image 3) of the backside of the vehicle (see fig. 8) and

sends a warning to the LX that the danger zone is blocked by a vehicle. • Camera 4 detects a vehicle with the expected value (see fig. 8) and sends a message to the LX that the

obstacle has left the danger zone and that the LX is free again. • Camera 3 detects the danger zone and makes a picture at t1 (image 5). If there is no difference between

image 2 and image 5, the danger zone is free of obstacles. • The barriers can be closed.



PROCEEDINGS 183/208

Fig.9. Sequence of obstacle detection in the LX danger zone

If the system is not able to generate a doubtless vacancy proving detection of the danger zone, the LX will be signalled as not secured. By this a misleadingly as free signalled LX can be avoided.

Demonstration

Because of a wide operational area of such a method, it is necessary to perform realistic tests. Especially with regards to the safety criticality of such an application, first tests will be done in a non-public area where only a SIL 2 (SIL = “Safety Integrity Level”) system is required. For the field tests, a road-rail vehicle and a minivan will be used in the first steps. After an initial phase of tests, a demonstration unit will be developed that can be mounted at a LX in the above mentioned non-public area.

Conclusion

The implementation of imaging methods using camera based technology can help increasing the safety of railways especially at level crossings. To implement such an innovative system, intensive test campaigns are necessary in which the multiple requirements regarding safety targets, availability, maintainability and security can be evaluated. Innovative systems using camera based technology form an economical advantageous alternative to existing track-fixed monitoring units still reaching the required safety regulations formulated by standard books, laws or other official documents all over Europe. The Institute of Transportation Systems of the German Aerospace Center in Braunschweig develops such a system and will evaluates it in different field tests. First results are expected in 2009

References

[1] Dieter Ellinghaus, Jürgen Steinbrecher: Das Kreuz mit dem Andreakreuz, Continental AG, Hannover, 2006 [2] Allgemeines Eisenbahn Gesetz (AEG) [3] Bernhard Mrochen: Zugselbstabfertigung durch den Triebwagenführer bei der Hamburger S-Bahn, SIGNAL

+ DRAHT, 4/2002 [4] Bruno Zürcher: Ungesicherte Bahnübergänge mit wenig Geld verbessern, www.Wochen-Zeitung.ch,

05.10.2006



PROCEEDINGS 184/208


Site Research, Simulation and Evaluation of Novice Engineering Safety Devices at Israeli Level Crossin gs

Author(s): Dr. Moshe Becker, Dr. Mina Zemach, Eng. Amos Gellert, Dr. Elia

Ben-Shabat, Eng.Yaron Ben-Dor

Job Title: Senior advisor To Israel Railways Ltd.

Company: Israel Railways Ltd.

Country: Israel Resume of Speaker Dr. Moshe Becker: Transportation and traffic safety consultant, advisor to Israel Railways Ltd. B.Sc. (1967) and M.Sc.(1971) in Civil Engineering, the Technion, Israel Institute of Technology. M.S.E. (1973) and Ph.D. (1975) in Traffic and Transportation Safety, the U. of Michigan, USA, 40 years of research, teaching and consulting experience in matters of Transportation and Traffic Safety including Railway safety. Member of the organising committee of ELCRF (European Level Crossing Research Forum).

Abstract

Following two major derailment accidents at level crossing in Israel, during the years 2005-2006, a technological committee was established to learn and recommend the implementation of various novice safety devices to improve level crossing (LC) safety. Among the directions taken was the implementation of novice engineering safety devices to reduce risks taken by motor vehicle drivers approaching an actively protected LC when a train is approaching. Also was considered a situation when motor vehicles accumulate downstream behind the LC to form a queue stranded on the LC. This paper describes the research methodology taken by the "Israel Railways Ltd" research team and presents as an example a summary of results related to the impact of one of the safety devices investigated, namely, LC Preceding Regular Traffic Lights installed to prevent vehicle drivers from committing LC violations or causing an accident risk when vehicles are stranded on the LC tracks.



PROCEEDINGS 185/208


The railway is not an island: Building partnerships with the wider community

Author(s): Sue Nelson

Job Title: Director

Company: Community Safety Partnerships Ltd

Country: United Kingdom Resume of Speaker

Sue Nelson is a former journalist and news editor of a British daily newspaper with more than 16 years’ experience in public relations, corporate reputation handling, crisis management and community relations. In recent years she has specialised in developing and delivering community safety messages and campaigns to cut crime and anti-social behaviour on Britain’s rail network. She is currently working with British Transport Police and rail industry to develop broader understanding of the external community safety agenda and develop capabilities in partnership working. In addition Sue has been personally involved in assisting those affected by major rail incidents and other events including the Southall, Ladbroke Grove, Hatfield, Great Heck and Potters Bar tragedies and now helps organisations consider how they would relate to those affected by disaster.

Scope

Risk arising from trespass, level crossing abuse and other crime and anti-social behaviour is not solely within the control of the railway. This paper sets out how railways can more effectively engage with wider civil society to reduce the harm arising from negative public behaviour and address fear of crime which leads to journeys not being made by train.

The issues at stake

The railway right of way has, since railways were first built, divided communities and so generated a tension between the railway authorities’ wish to restrict the points at which the public can cross the railway and those who wish to find the shortest route between two points. The consequence of this is on the one hand the responsibility to ensure that the railway can be crossed safely and on the other to address trespass. In each case, a range of engineered solutions have been deployed. However, the effectiveness of the engineered controls is a function of public acceptance of them and their resultant behaviour when faced with them.

Where a public right of way pre-dated the railway, the price of constructing the railway was generally the maintenance of the established right of way by means of a level crossing or bridge. However, changes in the built environment in the hundred and fifty or so years since many railways were built has created desire lines that are often not reflected in the siting of additional authorised crossing points, with the result that unofficial crossing points are established by elements of the surrounding community who consider “trespass” a right rather than an offence.



PROCEEDINGS 186/208

However, the issue of public behaviour is far wider than that of inappropriate crossing of the railway and needs to be considered in the context of the railway as a destination whether it be to travel by train, properly use railway facilities such as shops and restaurants available to the public, or to use the railway as a place to gather, play or commit crime against the railway itself or its staff and customers.

Historically, in Great Britain the down-side of public behaviour was referred to as trespass and vandalism, seen as something beyond the control of the railway and often trivialised. In so doing a further dimension was lost with little recognition that there are prospective passengers who either do not use the railway or use it in daylight hours only because they perceive that their personal security is threatened by crime and anti-social behaviour on and around the railway. Indeed, there are some who will think twice before using public transport at any time of the day because they feel intimidated by the dress and behaviour of others.

Given that the public behaviour seen on the railway is essentially the same to be found in the wider community, railway businesses operating independently and unilaterally or together will not tackle the root causes of negative behaviour and so are likely to displace rather than reduce the overall incidence of crime and anti-social behaviour in the community.

Rail industry arrangements

At the end of the 1990s the then rail safety regulator – the Health & Safety Executive (HSE) – along with its governing body – the Health & Safety Commission (HSC) – were instrumental in bringing the by then fragmented and privatised railway together in a national working group. This along with regulatory concerns that assaults on staff were not given appropriate consideration was the stimulus the industry needed to begin thinking of a broader range of crimes against the railway and those that work on it or use it. Where the industry was undertaking work to counter crimes against the railway, initiatives tended to be local and at times fragmented.

The rail industry concluded that it should take on responsibility for a national focus on crime against the railway and as a result established what was initially known as the National Trespass and Vandalism Group and subsequently the National Railway Crime Group, which addressed issues in the context of route crime (i.e. on the line of route), station crime and on-train crime. In parallel the National Level Crossing Safety Group and Rail Personal Security Group were established to provide a focal point for these issues.

By 2003 the term “route crime” had generally supplanted the long-established term trespass and vandalism or T & V for short. However, this term was not understood outside the industry and indeed inside the industry too. At this time there was an initial recognition that the industry needed to find a better way of describing the portfolio of public behaviour issues impacting on the railway and also that there was a case for more support for the industry’s national groups.

These factors came together in 2004 with the establishment of a Community Safety Support Unit funded as a research project by the Rail Safety and Standards Board (RSSB) reflecting the by then well established external term of community safety. By 2006 when the CSSU project came to an end, the structure of the national groups had changed with the establishment of a high level Community Safety Steering Group (CSSG) overseeing a network of Community Safety Partnership Groups (CSPGs) and the national groups providing a focus on level crossings and personal security with the support deemed necessary provided from within RSSB. To date, the effectiveness of the new industry arrangements has not been fully quantified, but it is clear there is a mixed response to the approach, with some elements of industry either not knowing how the CSSG/CSPG framework sits in the broader picture or acknowledging this as the way forward.

The CSSU project was significant in that it successfully raised consciousness that seeing the railway as an island wasn’t a good way to deal with issues that had their roots in the communities through which railways pass. It showed that the industry should therefore take account of the external agenda and arrangements for tackling the parallel portfolio of risks within society collectively known as community safety.



PROCEEDINGS 187/208

External agenda: the rise of community safety

Conceptually the seeds for managing community safety were first sown more than 20 years ago. A 1984 British government Home Office Circular25 decreed that crime prevention should be a significant and integral goal of national and local public policy. It stressed the need for a co-ordinated approach and joint strategies involving partnership. The Morgan Report26 from 1991 further developed the concept of ‘community safety’, emphasising that crime reduction should be ‘holistic’ covering both situational and social approaches. Morgan noted that crime reduction was a peripheral issue for major agencies and a core activity of none of them and advocated the development of multi-agency crime prevention initiatives co-ordinated by local authorities. Morgan identified six elements crucial to multi-agency crime reduction work: structure, leadership, information, identity, durability and resources. Although the term community safety achieved prominence from 1991 it remained a rather elusive concept that can be difficult to define in exact terms that are applicable to everyone. Therefore community safety is perhaps best seen as an aspect of ‘quality of life’ in which people, individually and collectively, are protected as far as possible from hazards or threats that result from the criminal or anti-social behaviour of others, and are equipped or helped to cope with those they do experience. It should enable them to pursue, and obtain the fullest benefits from, their social and economic lives without fear or hindrance from crime and disorder.

Importantly, priorities of local communities should drive the scope of community safety activities at a local level. The definition of community safety must therefore reflect the breadth of understanding in the wider community as it varies from place to place.

Community safety means more than the more commonly used terms crime reduction or crime prevention. In using community safety, it is recognised that the focus of attention should not only be on efforts to reduce or prevent crime and disorder, but also on securing social and economic change as a means of preventing crime and disorder. Community safety activities aim to reduce offending behaviour and also the harms experienced by individuals and communities because of crime and disorder, and will seek to improve their quality of life through efforts to change the wider physical and social environment. Furthermore, community safety deals with the fear of crime and perceptions of safety as, were these aspects not to be addressed, any initiatives to reduce crime and anti-social behaviour would be of lesser benefit.

An early response to the concept of community safety was the Safer Cities initiative launched in March 1988 by the Home Office as the flagship of the then government’s crime prevention policy. Phase 1 ran until 1995 with the aim of reducing fear of crime and creating safer environments for economic and community life to flourish. Safer Cities was locally based and took a partnership or multi-agency approach. In each of 20 cities or boroughs, the Home Office funded a project co-ordinator and a small team. Each team was guided by steering committees representing local government, the police, probation, voluntary bodies and commerce.

The projects tackled a range of crime problems, some focusing on the city as a whole, although most schemes concentrated on local issues. All were expected to adopt a problem solving approach by analysing data, adopting tailor-made preventative measures and evaluating the results. Railway businesses were engaged in some of the pioneering Safer Cities initiatives - for example, in the northern England city of Bradford station access arrangements were improved as a direct result of funding unlocked by the local Safer Cities partnership.

25

Home Office Circular 8/1984

26 Safer Communities: the Local Delivery of Crime Prevention through the Partnership Approach, 1991



PROCEEDINGS 188/208

This was probably the first formal interaction between the railway and an externally driven community safety initiative.

It was evaluation of this scheme that led to a 1997 Home Office consultation document27 which outlined proposals to legislate through what in 1998 became the Crime and Disorder Act. This document set out the government’s intent to maximise the contribution of key partners in crime prevention and community safety in a way that gave local people an opportunity to contribute to the improvement process.

It acknowledged and reiterated the importance of Morgan’s work and its assertion of the need for broadly based multi-agency approaches to crime prevention, and the need to involve both the voluntary and business sectors as partners. It noted that one of the biggest barriers to progress was the lack of a statutory role for local authorities who were to be the heart of the postulated approach. Businesses across the national rail system did not, at the time, recognise the potential benefits from engaging externally.

The consultation document made clear: “The proposals set out in this paper are not about requiring local government to deliver a major new service or take on substantial new burdens” rather “their aim is to give the vital work of preventing and reducing crime a new focus across a very wide range of local services, including… those provided by local authorities. It is a matter of putting crime and disorder decision-making where they have always belonged.”

External agenda: implementation

The 1998 Crime and Disorder Act placed a clear legal obligation on local authorities and the police jointly to develop and implement a strategy for tackling crime and disorder. Along with local fire authorities and primary health care bodies, the Act defined them as ‘responsible authorities’ who, when so acting, are required to involve a wide range of other key agencies, for example from the health, education, business, transport and voluntary sectors. They must also consult widely with the community as a whole. This approach recognises that both the causes of crime and disorder and the interventions required to deliver safer, more secure communities lies with a range of organisations, groups and individuals working in partnership. Crime reduction is not solely the responsibility of the police. The overall aim of sections 5 and 6 of the Act is to ensure that responsible authorities:

• Are aware of the nature of crime and disorder in their area • Are able to identify the methods of developing and implementing effective action to help reduce that

crime and disorder • Formulate and publish a crime and disorder reduction strategy setting out the findings from auditing the

levels and patterns of crime in their area and putting the strategy into practice It is clear, although many transport providers did not recognise this, that these overall aims are inclusive of transport systems within the geographic extent for which the designated responsible authorities are accountable. The responsible authorities’ awareness of local needs requires an understanding both of the nature of crime and disorder and its causes in order to develop an effective programme to help deal with it. The vehicle for this in England are a network of Crime & Disorder Reduction Partnerships (CDRPs)28 which undertake analysis of crime in their areas from which they identify the actions needing to be taken. Exploring how to develop the wider perspective of how crime impacts on the community and how the community can help impact on the crime has been a key element of this approach. In many cases the CDRPs have not sought to explore community safety issues on the railway, perhaps because of an insular attitude on the part of the rail industry and the fact that the railways are policed by a separate force, British Transport Police (BTP).

27

Getting to Grips with Crime: A New Framework for Local Action, Home Office 1997

28 Similar arrangements apply in Scotland and Wales



PROCEEDINGS 189/208

The view that tackling crime is a matter for the police, tackling poor health is a matter for the NHS, deprived neighbourhoods are a matter for local authorities and so tackling railway crime is a matter for the rail industry categorises every problem and isolates responsibility for dealing with it into the silo of an individual agency or organisation. This is not the intent of the Crime & Disorder Act of 1998 as it cuts across issues and adopts the central tenet of collaborative action to address shared problems. For people in the community the quality of their life in their neighbourhoods is affected by a whole range of influences. Therefore, local organisations working together can collectively provide interventions and responses to tackle problems and provide earlier, more effective solutions, some of which will be of value to the rail sector. The 1998 Act outlines in detail those bodies that should be part of the crime reduction process at local level. This quite clearly shows that the rail industry has a role to play in areas where there are crime and disorder issues, with the specific addition of transport providers to the mix being achieved through the Police Reform Act 2002. Yet five years after the specific referencing of transport providers in community safety legislation there appear to be a significant number of rail industry players who either do not understand the extension of scope to specifically include the railway or choose to ignore it.

External agenda: neighbourhood policing

In 2004 the British government set out a vision for forging a new relationship between the police and the public, a relationship that would rely on active collaboration between the police, their partners and communities in the delivery of policing services. Building on this the results of a trial of a National Reassurance Policing Programme (NRPP) at 16 sites in England between 2003 and 2005 which involved local communities identifying crime and disorder issues in their neighbourhoods which they then tackled together with the police and other public service providers and partners, found the results of the programme to be consistently positive. The evidence, published in 2006, showed that improvements in key outcomes such as levels of crime, perceptions of anti-social behaviour, feelings of safety after dark and public confidence in the police were directly attributable to this approach. The programme also delivered statistically significant improvements in trust among communities themselves. In parallel, the importance of neighbourhood and community was also recognised in the government’s public performance targets - Public Service Agreements (PSAs) for the Home Office, Department for Communities and Local Government, Department for Constitutional Affairs and the Crown Prosecution Service all reflect a commitment to neighbourhoods. This matches the commitment of the Association of Chief Police Officers (ACPO) to neighbourhood policing. Neighbourhood Policing, which from April 2008 is present in every community nationwide, is based on addressing local priorities identified by local people in their local area. At the heart of the Neighbourhood Policing approach are three guiding themes:

• To create permanent and dedicated teams that will have specific responsibility for each defined neighbourhood, and are familiar faces to those who live and work in that area

• The teams will use the National Intelligence Model to direct their activities - focusing on those problems that the public have told police matter most to them

• Neighbourhood teams will work closely, and take joint action, with local authorities, voluntary groups, businesses, criminal justice agencies and other partners to tackle these issues.

The Neighbourhood Policing Programme was adopted to support realisation of the government’s vision for policing that is accessible and responsive to citizens’ needs. Reducing crime levels and emerging levels of fear of crime has confirmed the value of the concept around signal crimes, reassurance policing, and community cohesion. In short, Neighbourhood Policing:

• Is an organisational strategy that allows the police, its partners and the public to work closely together to solve the problems of crime and disorder, improve neighbourhood conditions and feelings of security

• Is managed within mainstream policing activity, integrated with other policing services • Requires evidence based deployment of neighbourhood teams against identified need



PROCEEDINGS 190/208

• Establishes dedicated identifiable, accessible and responsive neighbourhood policing teams which provide all citizens with a named point of access

• Reflects local conditions and is flexible and adaptive • Allows the police to work directly with local people to identify problems that are most important to them,

thereby giving people direct influence over local policing priorities • Establishes a regime for engaging other agencies and the public in problem solving mechanisms • Uses the National Intelligence Model (NIM) as the basis for deployment • Requires an effective engagement, communication and feedback strategy, and a clear explanation of

where accountability lies • Should be subject to rigorous performance management including clear performance monitoring

against a local plan and commitments made to a neighbourhood

Rail sector CDRPs

The Neighbourhood Policing concept has been embraced by BTP who are also increasingly engaged within CDRPs in the wider community. As a result, elements of BTP have first hand experience of CDRPs, which they rightly consider applicable within the mainstream railway environment. On the national network BTP have in their London North Area set up five policing sector based rail CDRPs with membership drawn from a diverse range of rail industry stakeholders. It is too early to determine the longer term value of this initiative in terms of levels of and fear of crime. However, it is possible to identify that there is at local level a greater common ground between BTP and rail industry businesses. In other areas where police “tasking groups” have emerged with local rail business invited to participate, similar benefits are evident.

Greater progress in aligning the approach to addressing community safety on a transport network has been achieved by Transport for London (TfL) which has championed adoption of the CDRP approach that is integral to their first publicly available Community Safety Plan for 2007-2008.

TfL published a Draft Crime and Disorder Reduction Strategy in January 2007 setting out their intent to satisfy the generality of national community safety legislation. The subsequent publication of a Community Safety Plan can be seen as the point from which TfL will start of the process of reviewing policies and practice in the community safety arena. This review will in turn facilitate the annual updating of the community safety plan setting out objectives and activities to deliver them. The objectives and activities target both a reduction in crime and anti-social behaviour and the fear of crime. Going forward, the content of TfL’s future community safety plans will be shaped by the prior publication of a strategic assessment.

The rationale for the approach adopted by TfL can be tied back to their decision to voluntarily adopt Section 17 of the Crime and Disorder Act 1998, which requires authorities to consider crime and disorder reduction in the execution of all their duties. This means that TfL will, in the decisions it takes, consider their likely effect on crime, disorder and community safety. In turn this feeds through to doing all that it reasonably can to prevent crime and disorder, substance misuse and anti-social behaviour. In the current year TfL has committed to further extend the adoption of Section 17 of the Crime and Disorder Act 1998 across all areas of its organisation and are asking partners, including external CDRPs, to help assess the impact of the TfL approach to community safety.



PROCEEDINGS 191/208

Cutting Crime strategy and supporting National Comm unity Safety Plan (NCSP)

British crime statistics show that crime has fallen by around a third in the past ten years, but clearly new challenges continue to emerge. The social and economic context moves on and criminals innovate as quickly as those engaged in legitimate public and private enterprise. To account for this the British government has published Cutting Crime – A New Partnership 2008 – 2011 as the over-arching strategy for a wide range of plans and initiatives to tackle crime and community safety going forward. The document is also known generically as the Crime Strategy. Based on experience and evaluation of activity of the past decade in the context of emerging threats, the Crime Strategy identifies a number of priority areas to be addressed between 2008 and 2011:

� Stronger focus on serious violence � Continued pressure on anti-social behaviour � Renewed focus on young people � New national approach to designing out crime � Continuing to reduce re-offending � Greater sense of national partnership � Freeing up local partners, building public confidence

The new and further updated NCSP for the same period addresses what will be done to deliver the strategy. This plan is significant in that it provides for local partners having greater flexibility to deliver against their local priorities. The NCSP is for members of local partnerships with a role in delivering community safety. The Crime and Disorder Act 1998 obliges CDRPs to invite representatives of local businesses to become involved in the development and implementation of community safety strategies, and in addition, the Police Reform Act 2002 specifically identified the requirement for public transport providers and sponsors of public transport in metropolitan areas (Transport for London and provincially through Passenger Transport Executives) to be included. Rail businesses should therefore have the opportunity to input their own priorities and play a key role in shaping the direction of CDRPs covering localities identified by rail interests as a priority for action, so helping to ensure that tackling crimes against business plays an important role in their partnership plans. There is also a related statutory requirement for police authorities - including that for BTP - to annually produce a rolling three-year policing plan for their force area in which they set out their programme of action, bringing together national and local priorities. Clearly, rail business requirements are a key driver of this process. In considering how best to engage with those focused on meeting the challenges to be addressed within the context of the NCSP it is important for railway interests to recognise that, in the eyes of wider society, the railway is a homogeneous entity rather than a suite of component businesses. Therefore, it is important the rail industry has in place arrangements to ensure that engagement with CDRPs is cohesive and not one of dealing with a web of competing and divided interests. To do this the rail industry needs to agree who will lead on engagement with CDRPs and related entities and the role that each of the parties will play in delivering rail inputs to local partnership working designed to reduce crime, the fear of crime and anti-social behaviour. In many cases it is likely that BTP are best placed to lead on behalf of rail interests in the geographic areas that have been identified of greatest significance to rail. Remaining focused on the geographic areas and so the communities where the rail industry most needs to engage to address community safety risks is of fundamental importance. Spreading rail’s engagement too widely and therefore thinly is tantamount to abandoning the priorities for working through partnerships. However, whenever approached by a CDRP it is important that the rail industry evaluates the proposed interaction and provides reasoned feedback to the third party concerned. If rail interests simply ignore the needs of a host community’s priorities for action, the standing of rail in the community will be harmed.



PROCEEDINGS 192/208

Importantly, the Crime Strategy is supported by a new and radically simplified set of Public Service Agreements (PSAs) which set out the government’s objectives for public service delivery and explain how success in delivering these objectives will be measured. The new PSAs aim to demonstrate benefits of partnership more clearly at national, regional and local levels as they can only be delivered through multiple players working together. Many elements of the PSAs are directly relevant to rail as they address issues already identified in industry and company plans.

Implications for rail industry arrangements

The rapid evolution of the government’s arrangements to address the crime, disorder, fear of crime and anti-social behaviour agenda and the growing expectation that private sector organisations engage with the public and third sectors cannot be ignored. In particular, the rail industry needs to more generally accept that the providers of public transport have a particular role to play in addressing community safety risks. It is clear that an essential element of effective partnership working is high-level commitment within each of the partnering organisations, demonstrated by individuals being empowered to take strategic decisions and provide an oversight of both their own contributions and those of the partnership as a whole. Thus, there needs to be a strategy group which can:

� Commission strategic assessments (at least annually) � Agree a three year partnership plan � Commit the resources necessary to deliver what they have agreed shall be done � Monitor progress in delivery of the partnership plan (which needs to be refreshed annually)

In the context of the main line rail sector, this can be seen as a role that may best be taken forward by the Community Safety Steering Group (CSSG), which already exhibits a number of the characteristics evident in the arrangements put in place by TfL. However, the TfL approach evidences one very important characteristic that can be seen to go beyond the main line rail approach, as it has sought to integrate their arrangements for addressing community safety within the government’s national framework for addressing community safety risk. TfL have established the London Transport Crime and Disorder Group as a transport focused CDRP which interfaces with those in the surrounding community. As such, the TfL approach is transferable to the main line railway system and could be the role adopted by the CSSG. TfL have put in place arrangements through which they will be able to demonstrate their approach to community safety is aligned with and supportive of delivery of the Crime Strategy and supporting National Community Safety Plan for the same period. This follows through into TfL’s published annual community safety plan which shows there to be a cohesive programme with the potential to improve safety and security across London’s transport system through targeted activity to reduce crime and anti-social behaviour, and at the same time reduce both public and staff fear of crime. The expectation that transport providers participate in community based partnerships cannot be avoided. Therefore, the national rail sector will need to be seen to have in place arrangements that take account both of a continuing belief that the railway is a single entity and the reality that railway services are delivered through the collaborative efforts of many.

It is important that rail industry arrangements for local delivery combine the energies of the many to present the rail sector as a whole wherever this is appropriate. Notwithstanding this approach, there will be many occasions where rail industry partners agree that one of them will take the lead. Therefore a primary role of locally based industry partnership working is tasking. Tasking groups have emerged in parallel with the industry’s second tier structure of Community Safety Partnership Groups and it is now appropriate for the main line rail sector to review the efficacy of these parallel approaches and determine how best to move forward.



PROCEEDINGS 193/208

Added value

National crime statistics and other studies show adoption of a broadly based community safety approach can reduce crime, disorder and anti-social behavior. Given the parallels between the rail environment and those in the wider community, partnership working will, if further developed within the rail industry, most likely reduce harm, improve the travelling environment and reduce the fear of crime which has been assessed as reducing travel by some 10%.

Similarly, the industry should substantially enhancing engagement in targeted externally led partnerships, as this will contribute to securing progress in addressing community safety issues.

Engaging with external partnerships can also be a mechanism through which railway businesses can secure matching funding from public sources to increase the resources available to address community safety issues on the railway. Importantly, this includes actions to reduce the fear of crime which impacts on passenger revenues and creates a negative perception of railway businesses.

Greater engagement with host communities has wider benefits in that railway businesses which engage and are seen to effectively address community priorities, for example child safety, are less likely to have their reputation damaged significantly when harm arises in the host community.



PROCEEDINGS 194/208

Session 10 - Future Vision Chairman: Simon Fletcher, UIC, France

Australian Road/Rail Strategy – development of the joint national interface strategy in Australia

Author(s): Phil Sochon

Job Title: Deputy CEO

Company: Australasian Railway Association

Country: Australia

Resume of Speaker Phil has an extensive background in different modes of transport safety including operations management, policy development and implementation and strategic planning through various roles in the Royal Australian Navy and the Maritime Services Board – Waterways Authority. With expertise in occupational road safety (fleet safety) Phil has worked with the New South Wales Government in this field and his work was acknowledged as one of the most prominent safety projects aimed at improving driver and vehicle safety. Having commenced in his current position in 2004, Phil is managing industry-wide initiatives in operational and strategic rail safety issues. He is actively involved in development of the national disability standards policy and national model rail safety legislation and the subsequent rollout of legislation all jurisdictions in Australia. Phil is also a Program Chair for Operations and Safety Research in the Australian Rail Industry’s Co-Operative Research Council. In relation to level crossings activities, Phil leads the rail industry involvement in the national joint Government and Industry behavioural Strategy which has successfully functioned for two years to date. Furthermore, he leads the Industry’s level crossing strategy implementation.



PROCEEDINGS 195/208

Session 10 - Future Vision

Applying a partnership approach to level crossing r isk: a strategic opportunity

Author(s): Aidan Nelson

Job Title: Director

Company: Community Safety Partnerships Ltd


Resume of Speaker Aidan Nelson has more than thirty five years railway experience of which eleven in the field of safety built on an operational career which included director level posts in both British Rail and the successor infrastructure manager. In April 2007 Aidan joined Community Safety Partnerships Ltd and is a director of the company. In recent years Aidan has facilitated development and delivery of health, safety and environmental policy both internationally and nationally at industry, multi-party programme and individual company levels. Aidan’s approach to level crossing and other community safety issues is to focus on the development of an enabling framework within which improved performance is secured through a balanced approach to engineering, education and enforcement initiatives. Abstract/Scope

Effective management of the risk arising at level crossings is far more than the railway progressively upgrading the technical integrity of the crossing and must encompass consideration of wider highway safety issues and, importantly, public attitudes towards and behaviours at level crossings. This paper, albeit from what may be seen as a rail sector perspective, identifies a strategic partnership based agenda for action.

Differential modal significance: a driver of divergent attitudes

As railways address the risks most directly within their control, collision with road vehicles on level crossing takes on a greater significance within the precursors of catastrophic rail accidents. Indeed, in many jurisdictions risk arising at level crossings is now right at the top of the rail sector’s priorities for action.

But, too often, it is said, that the rail sector has sole responsibility for managing the risk arising at road – rail intersections. This is illogical – even where a highway existed before the railway was built a hundred years or more ago as in reality both they and their use have changed greatly since then. Understanding divergent attitudes is helpful in determining how best to tackle risk arising at the modal interface. From a railway perspective there is significant risk from infrequent but catastrophic “train” accidents – in the United Kingdom most recently at Ufton Nervet in November 2004 – whereas from the highways perspective, collisions on level crossings are a very small component of “road” accident risk. Within the UK, level crossings typically account for some 10 fatalities per annum of which the majority are pedestrians which sit alongside the more than 3,000 occurring on the roads.



PROCEEDINGS 196/208

At both an institutional level and within society at large, there is an abhorrence of catastrophic rail accidents sitting behind the rail sector’s track record across the developed economies of long term reductions in risk within their direct control. It is of similar note that long-run improvements in road safety are also being secured. Indeed, in some jurisdictions, these improvements – many of which relate to survivability built into vehicles – have occurred at a faster rate than the generality of rail safety However at level crossings, recent UK practice has not delivered this sustained reduction in risk from both road and rail perspectives. The relative infrequency of fatal “road” accidents at level crossings when combined with a societal abhorrence of fatal railway accidents explains the high profile accorded by the media to these events. Indeed, a fatal accident on a level crossing whether the consequences are to road vehicle occupants can make the national news media whereas a similar accident at a signalised highway intersection may only generate localised and low-key reporting. Thus, if sustained improvements in level crossing safety are to be secured, there has to be a reconciliation of the differing modal perspectives. Key to this reconciliation will be an agreement of the roles and responsibilities of these and other players. Funding has the potential to be a sticking point but given the essentially public funding mechanisms for both rail and road infrastructure in Europe this will be less of an issue if there is political support for a shared approach using rational and shared decision making criteria than is likely when the rail infrastructure is or is considered to be privately owned. Capital expenditure: a matter of public policy Universally, there is a governmental role in setting the “rules of the game” nationally (or within the European Union multi-nationally) to govern the criteria to be applied when determining how and where public funding of infrastructure improvements is to be justified. In Sweden, for example, great progress has been made in developing a model in which there is an equality of treatment of road and rail infrastructure enhancement. Importantly, the Swedish model allows delay saved on the highway to be valued and used within the justification for the elimination of a level crossing. Simplistically, this philosophy is most likely to satisfy the railway perspective that the best level crossing is a closed one and road users who see level crossings as a significant source of delay. Equally important is the Swedish adoption of the “Vision Zero” goal that it is unacceptable for anyone to be killed or injured for life in a road accident, including those at level crossings. This when linked to application of the multi-agency partnership approach integral to the Swedish National Road Administration’s (Vägverket’s) accident prevention model OLA at level crossings29 shows road and rail modes acting in concert and working to address a shared risk through an agreed suite of actions for delivery by each party as is appropriate. In Great Britain there is a supposed equality when it comes to valuing safety improvements on both road and rail modes. However the value of preventing a fatality (VPF) updated annually by the Department for Transport (DfT) is in practice applied differentially by road and rail with the latter more likely to spend up to the national value30 used within quantitative analysis to support decision making. However, reconciling differences of national approach at the road – rail interface to reach agreement on physical changes at the modal interface is possible as was evidenced after the Great Heck collision31 between a train and a road vehicle, where the road vehicle had run off the highway onto the railway at a road–over–rail bridge. Here the DfT took account of recommendations made by working groups constituted by the regulator responsible for railway safety and the Highways Agency and published guidance and protocols – “Managing

29

http://www.vv.se/templates/page3____19602.aspx

30 RSSB has, using DfT guidance calculated the 2008 VPF to be £1.652m with 10 major and 200 reportable / 1,000 non-reportable minor injuries

equating to a single fatality.

31 February 28

th, 2001 – 10 rail staff / passengers were killed



PROCEEDINGS 197/208

the accidental obstruction of the railway by road vehicles32” – in February 2003. This document addresses the process of managing road - rail interfaces, responsibilities and cost sharing arrangements along with mechanisms for ranking risk. This document is of significance as it was developed with input of and agreement between representatives from national and local governmental perspectives and thus all public authorities with highways authority status as well as the rail sector. However, it is perhaps of greater significance in that it specifically excludes the road – rail interface as manifest at level crossings. Why? Because, at the time, this was a step to far given the historic national approach that saw level crossings to be a component of rail rather than road and rail infrastructure. Subsequently, Network Rail – the principal rail infrastructure manager in Great Britain – led the charge to convince parliamentarians that there should be a statutory obligation on highways authorities to participate in the management of risks at level crossings. This has been achieved within the Road Safety Act of 2006 which provides for statutory level crossing orders being used as a vehicle through which obligations can be placed on a highways authority, albeit at the expense of the rail infrastructure manager. Looking at these issues from a third perspective, that of the Netherlands, where the wider question of risk at level crossings was considered by the erstwhile Dutch Transport Safety Board 33 when investigating an individual accident at Voorst in 2000. This investigation34 highlighted that the Dutch Government had funded a substantial level crossing risk reduction and set out their policy government’s policy regarding the improvement of safety on level crossings: "The objectives for safety at railway crossings were formulated in the Framework Policy Document on Rail Safety (Lower House, meeting year 1998-1999, 26 699, no. 1). In the implementation of the policy the pursuit of optimisation is unremitting. The safety on level crossings is the result of the effort of the manager of the road and the effort of Railinfrabeheer [now ProRail]. A mutually agreeable approach in particular will have much effect in this matter. The interests of Railinfrabeheer and the road authority are largely similar." However, the Dutch Transport Safety Board stated that “In essence, this policy can be described as ‘vague’.” This was because rail and road authorities were found to have equal status and a split responsibility for rail and road components of the level crossing systems. In the absence of any inter-modal agreement concerning the holistic management of level crossing risk and transparency of the arrangements for cost attribution, progress in reducing risk was found to be sub-optimal. Taken together, the above European perspectives highlight that progress in managing the risk arising at level crossings can be achieved by reconciling differing modal perspectives and that the “rules of the game” concerning responsibilities – including funding – are a matter of public policy that cannot be ignored. In short, governments are best placed to put in place the high level enabling framework around which road and rail infrastructure managers can together build relationships with other institutional and community stakeholders to address risk arising at the road – rail interface. Added value The social costs of accidents at level crossings; those of providing, maintaining and operating the crossing itself; the capacity of the highway and that of the railway; and, the costs of delay to road and rail users all need to be considered together. This consideration will add value because the agreed decisions of the coalition will take account of the total costs and benefits of changing configuration or applying any other control. Thus, sub-optimisation based on the principle of “someone else will pay” should reduce. Who should take the lead?

32

www.dft.gov.uk/pgr/roads/network/policy/obstructionrailways/manaccidentalobstruc3909.pdf

33 Rolled into the more broadly based Dutch Safety Board in 2005

34www.era.europa.eu/public/Documents/Safety/Investigation_reports/NL/VOORST_NW%20definitief%20_januari%202003__engels%20_incl%20bijlag

en_EN.pdf



PROCEEDINGS 198/208

Although it is clear that there is a philosophical equality between road and rail modes in delivering safe level crossings, it is equally clear that there has to be one party which is the catalyst for action. In many cases the context is one in which there are multiple rail infrastructure managers within a single jurisdiction and therefore the potential for a fragmented approach were each to seek to lead independent of the others. Even greater numbers of highways authorities in a single jurisdiction are the norm. In order to avoid the trap of divergent and differentiated approaches it is important that a facilitative champion for action within the context of an over-arching national policy is found. Examples of such facilitative champions are to be found in various guises but mostly from a rail sector perspective where a dominant infrastructure manager or rail industry association. Given the significant costs associated with level crossings, there are those who believe that all the railway is seeking to achieve is a transfer of costs to others and promoting education to avoid the need to upgrade level crossings. From the author’s perspective there is more than a grain of truth in such reactions given the hectoring and preaching that can characterise the approach of some in the rail sector who feel that others are not playing a part. It is possible for a rail sector organisation to develop and provide effective facilitation for a group representing a wide range of stakeholders in level crossings. For this to work there has to be a separation of roles within the rail representation on the leadership group so that the convening and management of the group is inherently neutral and discrete from the commitment within a rail infrastructure manager to take action of a particular nature. Examples of rail industry facilitation of a national level crossings group can be found in Great Britain through the Rail Safety & Standards Board and in Australia where the Australasian Railway Association has taken on the role of championing change in user behaviours as a key strand of initiatives to reduce risk at level crossings. A further degree of separation and neutrality is to be found in North America through Operation Lifesaver which since small beginnings in Idaho in 1972 is the umbrella organisation through which initiatives, particularly those which are community based with a focus on education of road users, are driven. Significantly, there is direct government grant funding of a proportion of Operation Lifesaver costs which is analogous with initiatives to improve the generality of road user education. Indeed, it is self evident that educating road users in the safe use of level crossings is an essential strand in developing the competence of road users be they vehicular or pedestrian. Recognising level crossings as a feature of the highway network that should be addressed in road user training is an important step on the way towards driving tests – at least in respect of theoretical knowledge – covering their safe use. Initiatives such as Operation Lifesaver will only be sustained in the longer term when it is recognised that education of road users is not the only thing needed to address risk arising at level crossings. Clearly, there are reasonably practicable opportunities to reduce risk through engineering of the crossings themselves which is within the gift of rail and highways infrastructure managers working together. Indeed, commitment to and evidence of achieving good practice installations should be the bread and butter of these infrastructure managers. However, there efforts can at times be compromised by the relevant planning authorities giving inadequate consideration to the impact of their decisions on permitted development on level crossings. Operation Lifesaver has for many years adopted the mantra of the three Es: Engineering, Education and Enforcement to which a fourth strand for Enabling was formalised and added by the British in 2002. The 5th E – which is too often silent – is that for Evaluation Beyond Operation Lifesaver As set out above, Enabling is putting in place mechanisms that help a diverse range of stakeholders come together to share experiences, develop a shared understanding of risk at level crossings and agree actions that each will lead and where they will contribute.



PROCEEDINGS 199/208

A coalition of the willing: Railways, highways and planning authorities, Road vehicle and driver licensing authorities, equipment suppliers, motoring associations, commercial operators of road vehicles, volunteers from the community, law enforcement agencies, and of course representatives of the road user community: professional bus and truck drivers, amateur car drivers, farmers, pedestrians

Engineering is at the heart of the rail contribution to reducing risk at level crossings with many administrations working to secure an upgrade from a passive to an active crossing or, where risk is higher, from an active crossing protected by lights and bells to one with the added protection of half barriers. The cost of delivering these upgrades is never insignificant and sometimes prohibitive given the likely risk reduction. This cost dilemma sits behind research35 by the Rail Safety & Standards Board (RSSB) which highlights very significant cost differences within Europe and between European countries and Australia and North America. Benchmarking is an effective way of identifying tried and tested approaches which might be transferred to other jurisdictions. This isn’t just about new and novel approaches and is equally relevant when looking at the efficiency with which solutions are rolled out elsewhere. This latter dynamic is at the heart of the RSSB study highlighting up to a seven-fold difference in the cost of apparently analogous active crossings. This yawning gap between different administrations offers many an opportunity to learn and do more within a finite budget. However, on its own, this is not enough as the real wins will come when selectively upgrading level crossings for which closure cannot be justified is paralleled by targeted education of road users. However, engineered solutions are not the preserve of railways alone as highways authorities can do much too, as it is on their infrastructure that road users take their decision to cross the railway. Enhanced pavement markings, advance warning signs and traffic calming all have a part to play. Median strips to prevent zig-zagging are a case in point and it is incumbent on national highways standards to provide for these and other measures to reduce risk. Education is always going to be necessary as the way in which road users think and behave lies at the heart of level crossing risk. “Operation Lifesaver” has a track record of sufficient length to allow comparison of the relative efficiency of education and engineering initiatives. Indeed, an academic study36 shows education to be particularly cost-effective. In this context, there are early signs in Great Britain that Network Rail’s “Don’t Run the Risk” campaign, launched in the summer of 2006, is of value. Enforcement needs to be deployed alongside education to address those who persist in not using level crossings properly. Again, where photo-enforcement is used sensitively, technology has a real part to play. Evaluation of the benefits gained from each action to address risk is of paramount importance as the large numbers of level crossings found around the world mean that action must be targeted as few are in the fortunate position of achieving an environment in which grade separation of road and rail is universal. Learning from accidents and their precursors is a core component of evaluation. However, too many investigations by national investigative bodies and rail authorities have historically only really considered the railway engineered solution and can be seen as driving a spiral of upgrades which wouldn’t occur if equal effort was afforded to consideration of highway configuration, planning issues and road user behaviour and underlying competence. One of the issues associated with the accident investigation processes and their role in promoting learning from level crossing accidents is that too often the road user is killed in the collision. Therefore, first hand input from the road user perspective isn’t readily available if only fatal accidents are investigated. To make up for this, consideration needs to be given to a deeper investigation of precursors and events which have neither killed

35

www.rssb.co.uk/research/index.asp - T364: The cost of level crossings - an international benchmarking exercise

36 www.oli.org/savage_report.html



PROCEEDINGS 200/208

the road vehicle occupants nor derailed the train. Forward facing cameras on trains have the potential to be both a source of intelligence post-accident and importantly as a leading indicator of locations at which there are problems. This latter opportunity will only be realised when the costs of assessing the images to identify road-user risk taking have been automated and dramatically reduced in cost. Beyond this, there is a case for systematic covert surveillance of known bad-actor crossings and a control of similar crossings without a substantive accident / near-miss history to identify patterns of inappropriate behaviour and thus consideration of measures that might be taken to address risk at both individual level crossings and more generally at a type of crossing. Importantly, such measures may be engineered, educational or enforcement in nature or any combination of the three. International problem, international resources The same issues are found around the world and there is a great opportunity for shared learning and shared solutions. Nationally and increasingly internationally conferences at which level crossings are discussed are legion. There are workshops around the world at which the same ideas for research are put forward and outputs of research and other experiences considered. However, the knowledge that is contained within conference papers and research reports is fragmented and held in a multitude of different places. There are pockets of long-standing international collaboration, for example between the United States and Canada and more recently the research related protocol between the US and the United Kingdom37. Also, as an output of the 2004 Level Crossing and Trespass Prevention Symposium there is an European Level Crossing Research Forum facilitated from the UK and currently chaired by a German highways professional. The idea of a national coalition of the willing as a more effective means of reducing risk at level crossings than leaving it to the rail authorities alone is increasingly accepted. However, what is needed is a similar international coalition able to stimulate a greater sharing of experience and collaboration going forward. The concept of international collaboration needs to go beyond researcher talking to researcher and address the real issues that players are today tackling in parallel. For example; second train coming, access for persons of reduced mobility and the cost of parallel national equipment approvals regimes. Taking the question of approvals as an example; these are essentially national and anyone wishing to supply internationally is faced with the task of obtaining multiple approvals. There is thus a strategic opportunity for an improvement in the efficiency with which level crossing upgrades are delivered by applying the principles of aircraft type approvals which are essentially accepted on a worldwide basis; why can’t it be the same for level crossings? Just think of the economies of scale and more competitive market that would result. Maybe, creating a truly world market for level crossing equipment is a step too far because it involves a host of national standards and practices that many will seek to defend by focussing on a few relatively insignificant differences as an antidote to a focus on the far greater commonality of approach and philosophy for the control of risk so far as is reasonably practicable. If the approvals barrier cannot be cracked easily, then there should be a single web-based resource where an international coalition of the willing choose to share their knowledge with others working to reduce the harm that can arise on level crossings whether in the form of a very infrequent catastrophic train accident or the multiple accidents each of which proves equally devastating to the families of road users that die or are injured on level crossings around the world.

37

Protocols setting out a basis of cooperation are in place between the Federal Railroad Administration (FRA) and Transport Canada and between The

FRA and the Department for Transport (as successors to the Strategic Rail Authority) on behalf of the Rail Safety & Standards Board.



PROCEEDINGS 201/208

In conclusion

The recommended partnership approach – both nationally and internationally – will deliver value as partners accept that the management of risks at level crossings is not the responsibility of the railway alone. Once this principle is accepted coalitions of the willing are better placed to find optimal solutions which take full account of the benefits to be gained from highways enhancement, level crossing elimination, user education and enforcement alongside the traditional focus on the railway upgrading the configuration of the level crossing itself.

World-wide, many of the controls to address level crossing configuration, including on the approaches beyond the railway boundary are known. Likewise there is a large corpus of knowledge surrounding opportunities to lower the cost of providing enhanced level crossings. Similarly, there is a wealth of experience in changing attitudes towards particular risks through education and the role of enforcement within community safety strategies. Importantly, there are good precedents in road and rail working together and community engagement in education. Given these precedents and the available corpus of knowledge, it is clear that having agreed objective decision making processes within which the relevant authorities and the local communities affected by level crossings work as partners to determine the most appropriate measures to reduce risk is a realistic way forward on an evolutionary rather than revolutionary basis.

Further value of the broadly based coalition approach to managing the risk arising at level crossings is that it should encourage innovation and stimulate the quest for lower cost solutions. It is also worth considering the positive value of the various agencies and host communities being seen to work together in terms of damaged reputations, particularly of the railway, when it can be evidenced that risk has been managed so far as is reasonably practicable.



PROCEEDINGS 202/208

Session 10 - Future Vision

Moving forward together

Author(s): Mr. Simon Fletcher

Job Title: Senior Safety and Interoperability Manager

Company: International Union of Railways (UIC)

Country: Mother country, the UK, country of residence, France; country in

which office is based, Belgium Resume of Speaker

Since September 2003, Simon Fletcher has been coordinating the safety and operations activity at the UIC (the International Union of Rail Companies) located in Paris. On behalf of the railway community and working closely with the European Commission, he led the operations sub-system elements associated with implementation of the European directive on rail system interoperability. In 2004 the focus started to shift towards the European Rail Safety Directive and forging links with the European Rail Agency (ERA). Simon has been closely involved in development and implementation of the Safety Directive and the safety management system that train operators will be required to prepare. He is now based in Brussels, the home of the European Commission, as the UIC’s representative there and in a good position for operations and safety issues Much of this activity has been conducted through the medium of the UIC’s Safety Platform that works in several dimensions, amongst which are safety system development and the management of safety at the operational interfaces; the latter activity which he leads. Coming from an operations background within the UK, he has held senior safety and operational roles as well as having experience in operational rules and standards-setting and with development of the Eurostar service through the Channel Tunnel. In his earlier career he was a train driver as well as a driver tutor.

Abstract

The presentation draws on the highlights from the three days of the Symposium, recall the core themes, and marshal the principal messages into some forward-focussing options for future development.



PROCEEDINGS 203/208

Technical Visit

Summary:

� extracts from the French Highway Code;

� presentation of the level crossing visited;

� presentation of the planned removal operation.

Highway Code

Extracts from articles relating to level crossings

Article R412-29. The traffic lights for vehicles are green, amber or red. The red and amber traffic lights may be flashing.

Article R412-30.

Drivers must stop at a red light, whether flashing or not. You must stop behind a line that is perpendicular to the axis of the road. When this stop line is not present on the road, you must stop level with the traffic lights or before the pedestrian crossing if there is one. Any driver who contravenes the provisions of this article will incur the fine applicable to fourth class contraventions. Anyone guilty of this infraction will also incur the additional penalty of disqualification from driving, for three years at the most; this disqualification may be limited to driving excluding professional activity. This fine will automatically incur four penalty points on the driving licence.

Article R412-41.

When the crossing of a railway line is controlled by flashing red lights, it is forbidden for pedestrians to cross this railway line at any time whilst the light is flashing.

Article R412-43

Any pedestrian who contravenes the provisions of this section will incur the fine applicable to first class contraventions.



PROCEEDINGS 204/208

Article R414-12. Overtaking is forbidden when crossing railway lines not equipped with barriers or half-barriers. Any driver who contravenes the provisions of this article will incur the fine applicable to fourth class contraventions.

Article R416-19.

When a vehicle stopped on the road constitutes a danger for the traffic, especially near crossroads, bends, the brow of a hill, level crossings and when visibility is poor, or when all or part of its load falls on the road and cannot be cleared immediately, the driver must ensure advanced signalling of the obstacle by using the hazard lights or a signalling triangle or both.

A bylaw from the Ministry of Transport sets the application conditions for this article and determines the statutory equipment for each category of vehicle.

Any driver failing to use advanced signalling of a vehicle or obstacle within the conditions set out in this article or those taken for its application will incur the fine for fourth class contraventions.

Article R417-9.

Any vehicle stopped or parked must be positioned in such a manner as to not constitute a danger for road users.

Stopping or parking near crossroads, bends, the brow of a hill and level crossings, when visibility is insufficient, are especially considered as dangerous.

Any dangerous stopping or parking will incur the fine for fourth class contraventions.

When the driver or holder of the registration papers is absent or refuses, in spite of the officers’ order, to move from the dangerous parking site, immobilisation and taking to the pound can be enforced within the conditions provided for in articles L. 325-1 to L. 325-3.

Any driver guilty of one of the infractions stipulated in this article may also incur the additional sanction of disqualification from driving for three years at the most; this disqualification may be limited to driving excluding professional activity.

This fine will automatically incur three penalty points on the driving licence.

Article R422-3.

I. – - When a railway line is laid on a road or crosses it on a level, the priority is to the machines normally circulating on this railway line, with the exception of public transport vehicles obliged to follow permanently a fixed route determined by one or more rails and using the bed of the road, whose drivers must respect the road signs giving strict instructions and the indications given by traffic police. II. – No driver should enter a level crossing if his vehicle is liable to, because of its technical specifications or the traffic conditions, be stopped on it. When a level crossing is fitted with barriers or half-barriers, no road user should cross it if the barriers are either down or being closed or opened. When a level crossing is not equipped with barriers, half-barriers or lights, no road user should cross it without making sure that no trains are approaching. When the level crossing is manned, the road user must obey the guard’s orders and not hinder, under any circumstances, the closing of the barriers.



PROCEEDINGS 205/208

III. – When a train is approaching, drivers must immediately drive clear of the railway line in order to let it pass. IV. – Animal herders must especially take any measure allowing them very quickly to stop their animals crossing the level crossing. V. – In the event of the forced immobilisation of a vehicle or herd, its driver must do everything in his power to clear the obstruction of the railway line as quickly as possible or, failing that, warn the railway personnel without delay of the existence of the danger. VI. - Any driver who contravenes the provisions of this article will incur the fine applicable for fourth class contraventions. VII. – Any driver guilty of one of the infractions stipulated in this article will also incur the additional sanction of disqualification from driving for three years at the most, this disqualification may be limited to driving excluding professional activity. This fine will automatically incur four penalty points on the driving licence.

Presentation of the Breuillet level crossing number 30

Type SAL4 level crossing: automatic signalling lights with 4 barriers; Railway line with 2 tracks, electrified (1500V), with a speed of 120km/h, used by 90 trains per day (RER C, TER); Departmental road number 26, in town, with 5,675 vehicles per day; i.e. a moment function of 510,750; 5 barrier breakages in 10 years; It is a level crossing that causes concern because of its configuration (complicated crossroads and proximity of the station platform); Reminder of the level crossing criteria causing concern:

� 3 collisions minimum in 10 years ;

� the moment function is higher than 1,000,000 ;

� the configuration presents a risk (proximity with a crossroads, short lock, …), according to the expert.



PROCEEDINGS 206/208

Presentation of the planned removal operation The level crossing in question is one of the 14 level crossings situated on the Dourdan branch of line C of the RER and therefore concerned by the launch of feasibility studies for their removal planned within the scope of the protocol related to the emergency plan for the improvement of the regularity of RER D and the realisation of feasibility studies on line C.

A feasibility study was already done in 1997 by the Conseil Général de l’Essonne which suggested the crossing of the tracks east of the level crossing by an underground passage for small-sized vehicles, heavy goods vehicles being diverted towards the route to be developed on the RD 82 within the scope of the removal of level crossing 28. In a letter addressed to RFF in May 2005, the Mayor points out that the removal of level crossing 30 is a priority today for the commune and that he is in favour of this solution.

The removal is planned with that of level crossing 28 situated about 500 m away; Elements arising from the functional analysis: The present level crossings 28 and 30 are used for:

� non-motorised traffic (bikes, pedestrians, especially for level crossing 30 because there is an access to the platforms of the Breuillet-Village station)

� private cars � tractors � heavy goods vehicles.

These level crossings ensure the link between the old town centre of Breuillet and the Port Sud quarter. They also give direct access from the RN20 towards the centre of Breuillet and localities situated in the Vallée de l’Orge, and lead to agricultural land.



PROCEEDINGS 207/208

The removal operations aim at making safer the crossing of railway tracks used daily by about a hundred trains (RERC, main lines trains to Châteaudun and Tours) and to divert the heavy goods vehicles traffic in transit which cross part of the town of Breuillet. The programme of removal of level crossing 28 consists in the creation of a road bridge linking RD116 to RD19, whereas the removal of level crossing 30 will result in the crossing of the tracks by small vehicles by means of a rail bridge. Physical and natural environment: The local topography cut by the valleys of the Orge, la Renarde and la Rémarde results in a climb to the plateau with steep slopes; the valleys are damp and flood plains. The water tables are near the surface in certain places, the ground is unstable (north-west sides of Breuillet). The environment includes sensitive natural spaces (wooded spaces and damp environments), with 2 natural zones of ecological, wildlife and flora interest, as well as animal and vegetal species relying on water. Alternatives considered by the Commune: Several alternatives are being studied with the commune.



PROCEEDINGS 208/208