Vulnerable Road Users_adb()

Vulnerable Road Usersin the Asian and Pacific Region

ADB

Asian Development Bank

Vulnerable Road Usersin the Asian and Pacific Region

Asian Development Bank

The views expressed in this publication are those of the sectorconsultants concerned and do not necessarily reflect the views of theAsian Development Bank. The term “country” does not imply on the partof the Bank any judgment as to the legal or other status of any territorialentity.

Asian Development BankP.O. Box 7890980 Manila

Published by the Asian Development Bank

Printed in the Philippines

National Library of the Philippines CIP Data

Asian Development BankVulnerable Road Usersin the Asian and Pacific Region

ISBN 971-561-133-3 Stock number 010499

Photographs in this publication were supplied by Ross Silcock Ltd., UK.Cover photograph: Road project in Indonesia (ADB picture library).

This book was largely written and developed by Ross Silcock Ltd. in association with TransportResearch Laboratory (TRL) Overseas Centre, the United Kingdom (UK). It was produced under anAsian Development Bank (ADB) regional technical assistance (RETA 5620: Regional Initiatives in

Road Safety) that has also resulted in the production of a comprehensive set of guidelines for decisionmakers on road safety policy.

The RETA project was undertaken by the following team under Charles Melhuish, Senior Policy Specialist,ADB. Those asterisked (*) were the major contributors to this volume. The project team comprised AlanRoss* (project director), Mike Goodge*, Caroline Ghee, Chris Robson, Tim Selby*, and Kim Smith of RossSilcock Ltd.; and Amy Aeron-Thomas*, Chris Baguley, and Goff Jacobs of TRL Overseas Centre.

The project team would like to thank the 112 participants from 23 countries in the Asian and Pacific regionwho attended the ADB/United Nations Economic and Social Commission for Asia and the Pacific (UN/ESCAP) seminar held in Bangkok, 2-6 September 1996. These are listed in Section 9.

Support was also received from a steering group comprising representatives of funding agencies,international organizations, technical groups, and individual experts from developing countries. They are asfollows: John Flora of the World Bank, Alamgir Mojibul Hoque of Bangladesh, Ian Johnston of the RoadEngineering Association of Asia and Australasia (REAAA), Charles Melhuish of ADB, M. Rahmatullah ofUN/ESCAP, and C. Romer of the World Health Organization (WHO).

Thanks should go also to Graham Dwyer who edited the book and was responsible for much of its designwork and production.

Acknowledgments

Acronyms

ADB - Asian Development BankCARS - Conference on Asian Road SafetyCSIR - Centre for Scientific and Industrial ResearchCTEP - Calcutta Traffic Engineering Project (India)DITS - Dhaka Integrated Traffic Study (Bangladesh)Lao PDR - Lao People’s Democratic RepublicMAAP - Microcomputer Accident Analysis PackageNIEs - newly industrialized economiesNMHC - national major highways of the People’s

Republic of ChinaNMV - nonmotorized vehicleODA - Overseas Development Administration

(United Kingdom)PDMCs - Pacific developing member countries

(of the ADB)PRC - People’s Republic of ChinaRETA - regional technical assistanceSMV - slow moving vehicleTRL - Transport Research Laboratory

(United Kingdom)TRR - Transport Research RecordUK - United KingdomUNDP United Nations Development ProgrammeUN/ESCAP - United Nations Economic and Social

Commission for Asia and the PacificUS - United StatesVRU - vulnerable road user

Measurements

k m - kilometerkm/h - kilometer per hourm - metermph - miles per hourm/s - meters per secondped/m2 - pedestrians per square meterped/min/m - pedestrians per minute per meter

Conversions

1 mph = 1.6 km/h1 yard = 0.91 meter

VULNERABLE ROAD USERS

CONTENTSPage

1 INTRODUCTION ...........................................................................................................11 . 1 ADB Regional Technical Assistance Project ..........................................11 . 2 Structure ................................................................................................2

2 VULNERABLE ROAD USERS: AN OVERVIEW .......................................................32 . 1 Introduction ...........................................................................................32 . 2 Pedestrians ............................................................................................32 . 3 Nonmotorized Vehicles (NMVs) ............................................................32 . 4 Motorcycles ............................................................................................42 . 5 Data Overview........................................................................................4

3 PEDESTRIAN SAFETY ................................................................................................73 . 1 Introduction ...........................................................................................73 . 2 Pedestrian Accident Data ......................................................................73 . 3 Pedestrian Facilities ............................................................................ 1 23 . 4 Integrated Programs ........................................................................... 2 1

4 PEDAL CYCLES AND NONMOTORIZED VEHICLE SAFETY.............................. 2 34 . 1 Introduction ........................................................................................ 2 34 . 2 NMV Accident and Vehicle Data......................................................... 2 34 . 3 NMV Facilities ...................................................................................... 3 14 . 4 Integrated Programs ........................................................................... 3 9

5 MOTORCYCLES ........................................................................................................ 4 05 . 1 Introduction ........................................................................................ 4 05 . 2 Motorcycle Accident and Vehicle Data .............................................. 4 05 . 3 Motorcycle Facilities ........................................................................... 4 45 . 4 Integrated Programs ........................................................................... 4 7

6 RECENT DEVELOPMENTS ..................................................................................... 496 . 1 Introduction ........................................................................................ 4 96 . 2 Land Use Planning ............................................................................... 4 96 . 3 Road Hierarchy .................................................................................... 4 96 . 4 Design Standards ................................................................................ 5 06 . 5 Traffic Counts ..................................................................................... 5 16 . 6 Accident Data Systems ...................................................................... 5 16 . 7 Foreign Aid Projects ........................................................................... 5 26 . 8 Traffic Calming and Speed Reduction Measures ................................ 5 26 . 9 Safety Audit ........................................................................................ 5 3

7 CONCLUSIONS ......................................................................................................... 5 6

8 REFERENCES AND BIBLIOGRAPHY ...................................................................... 5 8

9 CONTACTS ............................................................................................................... 6 2


1 INTRODUCTION

This book on vulnerable road users (VRUs) has been produced as part of a regionaltechnical assistance project (RETA 5620: Regional Initiatives in Road Safety) underAsian Development Bank (ADB) Contract No. 95-367. A companion book, Road Safety

Guidelines for the Asian and Pacific Region, is intended as a source of reference and guidanceto the region’s senior decision makers with responsibility for road safety.

Analysis of available data shows that VRUs are involved in a high proportion of trafficaccidents in some countries of the Asian and Pacific region. A specific effort was made to tryto quantify the nature and scale of the road safety problems facing these road users and to

1.1 ADB Regional Technical Assistance Project

Plate 1.1 (above): Pedestrians exposed to risk inthe traffic stream.

Plate 1.3 (below): Typical cycle-rickshaw inBangladesh.

Plate 1.2 (right): Vehicles in conflict in thePeople’s Republic of China (PRC).

1


identify examples of good practice in provi-sion of safety facilities for this group. Thedefinition of vulnerable road users variesdepending on the context of a particularstudy. For the purpose of this volume, whichaddresses the needs of the Asian and Pacificregion, the main focus is on:

1. pedestrians;

2. pedal cyclists and othernonmotorized vehicles; and

3. motorcyclists.

This publication has two main objectives,as follows:

i. to try to assess the involvement of VRUgroups in Asia’s high accident rate; and

ii. to identify and summarize general de-sign considerations of VRU facilities inthe region and, in particular, to identifysuccessful innovative measures alreadyin use.

Although the main focus of this publicationis on current engineering practices, these meas-ures in isolation do not necessarily solve allVRU safety problems. Consequently, comple-mentary enforcement and educational measuresare also briefly discussed. Because of the lim-ited availability of detailed data, the publicationis not always able to explore specific casualtytypes within each VRU group, such as the veryyoung or very old, who may have a higher riskof being involved in accidents. The publicationfocuses most on pedestrian safety, aspedestrians are the road users at greatest riskin most countries, and considerable researchand information was available for this groupof VRUs. Serious efforts have been made toaddress pedestrian safety. By comparison, thereview of facilities available for nonmotorizedvehicles (NMVs) and motorcycles has beenlimited by the lack of attention given to them.

1 . 2 S t r u c t u r e

Following this introduction, Chapter 2provides a general overview of the problemsfaced by VRUs. Chapters 3 to 5 consider theroad safety situation for each VRU type andidentify appropriate measures to assist eachVRU.

The three main VRU modes discussed herehave different operating capabilities, needs, and

safety problems. Moreover, there has beenmuch more research into pedestrian safety thanNMV or motorcycle safety. Consequently, thechapters specific to VRU modes have taken dif-ferent approaches. The pedestrian safety chap-ter has been able to review remedial measuresspecific to pedestrians in the region while theNMV chapter, of necessity, is restricted to in-formation and observations that could be col-lected during the country visits, because littlepublished information exists on NMV facilities.

For each VRU group, the collected data ispresented according to six subregional group-ings for the Asian and Pacific region. These are:the newly industrialized economies (NIEs),Central Asia, Southeast Asia, South Asia,Central Asian republics, and Pacific developingmember countries (PDMCs). The countrieswithin each of the subregion groups are asfollows:

NIEsHong Kong, China; Republic of Korea;Singapore; and Taipei,China;

Central AsiaPRC and Mongolia;

Southeast AsiaBrunei, Cambodia, Indonesia, Lao People’sDemocratic Republic (Lao PDR),Malaysia, Myanmar, Philippines ,Thailand, and Viet Nam;

South AsiaAfghanistan, Bangladesh, Bhutan, India,Maldives, Nepal, Pakistan, and SriLanka;

Central Asian republicsArmenia, Azerbaijan, Kazakstan, KyrgyzRepublic, Russian Federation, Tajikistan,Turkmenistan, Uzbekistan;

PDMCs11 island nations including Fiji, PapuaNew Guinea, Samoa, and Tonga.

The countries shown in bold were visitedby members of the project team in order to col-lect data, observe traffic conditions, and holddiscussions with local safety specialists andresearchers. A developed countries group (Aus-tralia, Japan, and New Zealand) has beenincluded for comparison.

2


2.2 Pedestrians

In the 1970s, research conducted by theTransport Research Laboratory (TRL) of theUnited Kingdom (UK) identified pedestriansafety as a problem area for developingcountries. Pedestrians, in general, are involvedin more accidents in developing countries thanin developed countries. While this can bepartially explained by the low motorizationlevels in developing countries, traditionaltransport planning and traffic management havegenerally tended to focus on the safety ofvehicular traffic and have largely overlookedpedestrians. In addition, there tends to be ahigher percentage of pedestrian trips in thecities of the developing world due to theinaffordability of private transport and inad-equate public transport.

Local conditions do influence pedestriansafety, as can be seen by casualty age and vehi-cle types involved. Developing countries havealso been found, in general, to have a higherpercentage of child pedestrian deaths thanmotorized countries. While this may beexplained by the population distribution (devel-oping countries tend to have a much higherpercentage of children), remedial measures mustbe designed to reflect the target age group, itsrelative youth, and lack of maturity. Commer-

cial vehicles have also been found to have highrates of involvement in pedestrian accidents indeveloping countries. While this may be due totheir larger share of the vehicle fleet andincreased exposure, remedial measures mayneed to be designed specifically to targetprofessional drivers with little education or fi-nancial resources, and the transport operators andcommercial vehicle owners who employ them.

2.3 Nonmotorized Vehicles

NMVs such as pedal cycles, cycle-rick-shaws, animal carts, and handcarts have tradi-tionally been an essential means of transport-ing people and goods in the cities and townsthroughout the developing world, particularlyin Asia. In recent years, however, they havecome under an increasing threat that has led totheir almost complete disappearance in somemajor cities in Asia (for example: Calcutta,India; and Jakarta, Indonesia). A study com-missioned by the World Bank and carried outby PADECO Co., Ltd. in 1995 reports that thefactors threatening the future of NMVs in Asiancities are as follows:

a) increased motorization (including the in-creased use of motorcycles) and a con-

2 AN OVERVIEW

Measures to assist VRUs through road design and traffic control methods havebeen recommended at various international conferences on traffic safety in orderto recognize and reflect the special needs associated with these groups. Facilities

suggested usually include segregating VRUs from motorized traffic and improving the urbanenvironment through traffic calming techniques and pedestrianization. A recent suchinternational conference held in New Delhi, India, on 27-30 January 1991 resulted in the DelhiDeclaration on the safety of vulnerable road users. The recommendations and suggestions putforward in that and other such conferences are supported and expanded in this publication.

2.1 Introduction

3


sequent reduction in street space avail-able for safe NMV use;

b) increased trip lengths caused by changesin metropolitan spatial structure;

c) exclusion of NMV needs in urban trans-port planning and investment programs,resulting in inadequate facilities forNMVs;

d) a general trend towards modernization ofAsian cities that promotes attitudes thatNMVs are “backward”;

e) a tendency to believe that NMVs are thecause of urban congestion; and

f) excessive and often inappropriate regu-lation of NMV operation, including limi-tations on NMV ownership and total orpartial banning of their use.

Experience gained from visits to countrieswith high NMV usage during this RETA studyand reported upon by Goodge and Selby dur-ing a safety seminar sponsored by the ADB andthe United Nations Economic and SocialCommission for Asia and the Pacific (UN/ESCAP) seminar in Bangkok in 1996 tends toconfirm that these factors are operatingthroughout the region.

However, government roads and planningdepartments in a number of countries were ob-served to be providing, or are planning to pro-vide, facilities specifically for NMVs, or inconjunction with other vehicles that may bedefined as either slow or two-wheeled. Thesefacilities provide significant potential for re-ducing safety and capacity problems.

All NMVs can be considered to be amongthe group generally termed “vulnerable roadusers.” However, this report will show that inmany countries, their reported accident rateswere observed to be relatively low, particularlywhen compared to the statistics for motorcy-cles and pedestrians.

An exception to the relatively low level ofaccidents involving cyclists occurs in certaincities in the PRC, such as Beijing, wherecyclists are involved in about 70 percent of thetotal traffic accidents. In the PRC, cyclists gen-erally constitute the largest proportion of thetraffic flow in most cities.

Although data for the separate categories ofNMVs have been included for the purposes ofthis report, it is recognized that pedal cycles,cycle-rickshaws, and other NMVs all displaydiffering operating capabilities and thus have

different associated problems. So it should benoted that any facility should take into accountthe expected NMV traffic composition duringthe design stage.

2.4 Motorcycles

Motorcycles account for a large proportionof all motor vehicles in some of the countriesin the Asian and Pacific region. Due to theircomparatively low cost, motorcycles tend to bethe first affordable motor vehicles that can bepurchased and used by young drivers. Unfor-tunately, these riders have high-risk thresholds,limited training and testing, and, obviously, alack of experience, all of which contribute to-wards making them high-risk road users.

Unsurprisingly, accidents involving motor-cycles tend to be greatest where large motor-cycle fleets exist relative to the number of mo-tor vehicles. The continued growth of this vul-nerable motorized mode of transport may ac-celerate motorization and death rates in theregion.

2.5 Data Overview

The extent of the VRU safety problem isprimarily reviewed using data collected undera parallel UN/ESCAP study. However, datagathered from local officials during ADBRETA-related country visits, and from recentpublications, have also been consolidated andis presented along with the UN/ESCAP datawithin the individual VRU chapters. The datagathered for NMVs as part of the UN/ESCAPstudy concentrate on cycle-based two- andthree-wheeled vehicles.

Thus, pedal cycle and cycle-rickshaw datacannot generally be separated. However, fromalternative sources, the proportional split of theNMV types has been shown wherever possi-ble.

Accident data for the region are summarizedin Tables 2.1 and 2.2, which show that pedes-trians and motorcyclists are involved in themajority of fatal and injury traffic accidents.More than half of all fatal accidents in Malay-sia; Singapore ; and Taipei,China involvemotorcyclists, while in Hong Kong, China,more than two thirds of all fatal accidentsinvolve pedestrians.

From the data available to date, cyclists seemto be involved less frequently in reported

4


5

Table 2.1: Percent age of Fatal Accidents Involving Vulnerable Road Users (1993)

Subregion Country Pedestrian Cycle-based(percent) vehicle(1) (percent)

NIEs Hong Kong, China 67 4Republic of Korea 48 0Singapore 25 8Taipei,China 19 6

(2)

Central Asia People’s Republic of China 11(2)

9(2)

(1994)

Southeast Asia Malaysia (1994) 15(2)

6(2)

Thailand 9 4

South Asia Sri Lanka 45 17

PDMCs Fiji 43 3

Developed Australia 19 3Countries New Zealand 15 3

na Data not available.(1) Includes two- or three-wheeled cycle-based vehicles.(2) Data relate to deaths.Source: RETA project data.

Motorcycle(percent)

10 75164

10(2)

57(2)

na

16

1

na16

Table 2.2: Percentage of Injury Accidents Involving Vulnerable Road Users (1993)

Subregion Country Pedestrian Cycle-based(percent) vehicle(1) (percent)

NIEs Hong Kong, China 35 4Republic of Korea na 0

(2)

Singapore 17 5Taipei,China 13 5


na(1994)


5(2)


PDMCs Samoa 37 6


na Data not available.(1) Includes two- or three-wheeled cycle-based vehicles.(2) Data relate to casualties, excluding deaths.Source: RETA project data.

Motorcycle(percent)

17 6(2)

6962

74(2)

57(2)

26

9

na14


accidents than the other two groups of VRUs.Table 2.3 shows the split between vehicle

modes owned in selected cities throughout theregion.

Although there are high percentages ofmotorcycles in certain cities such as Phnom

Table 2.3: Percentage of Total Vehicles (Not Including Handcarts) by Vehicle Type inSelected Asian Cities

City (Country) Bicycle Cycle- Animal Bus Motor- Otherrickshaw cart cycle motor

vehicle

Dhaka (Bangladesh) (1990) 10.7 53.3 0.0 1.3 16.0 18.8

Phnom Penh * 47.1 4.2 0.2 0.0 43.6 4.9(Cambodia) *

Shanghai (People’s 95.9 1.4 0.0 0.3 0.5 1.9Republic of China) (1990)

Kanpur (India) (1992) 54.7 3.5 0.6 0.1 18.6 22.4

Surabaya (Indonesia) (1992) 40.1 4.6 0.0 0.3 38.7 16.3

Tokyo (Japan) (1992) 59.9 0.0 0.0 0.1 13.8 26.2

Penang (Malaysia) (1992) 49.6 0.6 0.0 0.0 29.0 20.8

Metro Manila * 12.6 0.7 0.0 0.7 8.4 77.6(Philippines)

Chiang Mai 4.2 0.6 0.0 0.0 72.3 22.9

(Thailand) (1992)

Hanoi (Viet Nam) (1992) 84.6 0.4 0.0 0.0 10.8 4.12

(*) Unknown year of source.

Source: PADECO Co., Ltd., 1995.

Penh, Cambodia; Surabaya, Indonesia; andChiang Mai, Thailand, it should be noted thatthere is still a high level of cycle ownership inmany cities, including Surabaya, and (with theexception of Dhaka, Bangladesh) they are themost widely owned NMVs in those cities.

6


3 PEDESTRIAN SAFETY

3.1 Introduction

P edestrian safety in the Asian and Pacific region is first summarized by accidentdata before engineering facilities for pedestrian safety are reviewed. As the provision offacilities alone cannot guarantee appropriate usage and improved safety, this chapter

closes with recommendations on how enforcement and education, and publicity campaigns canbe organized to complement safety engineering efforts and to reinforce safe road user behavior.

Walking accounts for a significant share of all urban trips in developed countries(i.e., about a third of urban trips in the UK). But the inability to afford private transport andunavailability of public transport result in walking representing a much greater share of urbanand rural trips in developing countries. While the lessons and experiences can be consideredfrom developed countries, pedestrian safety programs in the developing world must reflectboth the greater demand and the limited resources available to address pedestrian mobility andsafety. Accordingly, the measures discussed here will focus on low cost, and thus low risk,facilities that can be afforded on a wide scale and are easily implemented.

3.2 Pedestrian Accident Data

In comparison with other road user groups,pedestrians are involved in a significant numberof fatal and injury accidents in numerous coun-tries in the Asian and Pacific region (see Tables2.1 and 2.2).

Table 3.1 shows a comparison of thesepedestrian accidents for various countries inthe region. Overall, Hong Kong, China had the

highest share of pedestrian deaths (two thirdsof all fatal accidents), while pedestrians wereinvolved in 62 percent of the total fatal acci-dents reported in Dhaka, Bangladesh, 48 per-cent in the Republic of Korea, 45 percent in SriLanka, and 43 percent in Fiji.

Malaysia (a country with a large motorcyclefleet) reported pedestrians involved in more

Plate 3.1:Incompatible mix ofvehicles andpedestrians.

Plate 3.2 (far right):Pedestrians areoften exposed todanger duringconstruction work.

plate 3.1 plate 3.2

7


than 20 percent of total accidents, while Indiaand Thailand reported such an unlikely lowlevel of pedestrian accident involvement thatthe data should be double-checked.

Table 3.2 shows that the majority of pedes-trian casualties occurs among those of workingage. In Malaysia, older teenagers (age 16-20)were the second largest pedestrian casualtygroup compared to Fiji and Papua NewGuinea, where those under 15 were the sec-ond largest pedestrian casualty groups.

The pedestrian accident severity ratios,based upon the number of fatal accidents rela-tive to injury accidents, vary widely, as shownin Table 3.3.

While high ratios are found in the urbanizedeconomies of Hong Kong, China; and Singa-pore, the ratio for Dhaka, Bangladesh, was sus-

piciously low. India’s ratio was also lowcompared to Samoa and Sri Lanka. The typicalequivalent ratio in the UK is 46 pedestrian in-jury accidents reported for every one pedestrianfatal accident reported.

Too many countries are not yet able to evendetermine their pedestrian accident involve-ment rates.

Given these countries’ low motorizationlevels, vulnerable walk trips are likely todominate travel patterns, with pedestrians rep-resenting a large, if not the majority, share ofroad accident victims.

It is, therefore, highly likely that a signifi-cant amount of underreporting of VRU roadaccidents also occurs in developing countries.The severity ratios above imply that substantialnumbers of pedestrian injury accidents are not

8

Table 3.1: Pedestrian Accidents as Percentage of Total Fatal and Injury Accidents

Subregion Country Fatal accidents Injury accidents (percent) (percent)

NIEs Hong Kong, China 1993 67 35

Republic of Korea 1993 48(1)

na

Singapore 1993 25 17

Taipei,China 1993 19 13

Central Asia People’s Republic 1994 11(1)

naof China

Southeast Asia Malaysia 1994 15(1)

9(1)

Thailand 1993 9 na

South Asia Dhaka, Bangladesh 1996 62 32

India 1991 2 2

Nepal 1993 45(1)

16(1)

Karachi, Pakistan 1993 50(1)

na

Sri Lanka 1993 45 48

PDMCs Fiji 1993 43 na

Papua New Guinea 1992 33(1)

15(1)

Samoa 1993 50 37

na Data not available.(1) Data relate to casualties.Source: RETA project data.


being reported in the accident data systems.(However, some of the difference may be ex-plained by poorer emergency medical services,which would result in a higher proportion ofvictims dying in developing countries in com-parison to those in developed countries.)

Available statistics have been analyzed fora number of countries and recent trends inpedestrian casualties are summarized (bycountry) below. It should be noted that in somecases, data are available only on the percentagesof deaths while in other cases it is availableonly as a percentage of fatal accidents.

Although not directly comparable (a singletraffic accident could cause several deaths), thedata on percentage of fatalities or percentageof fatal accidents do give some idea of relativeimportance of pedestrian deaths.

a) NIEs

Hong Kong, China67 percent fatal accidents,35 percent injury accidents

In 1993, 336 road accidents involving deathsand more than 15,000 road injury accidentsoccurred in Hong Kong, China, with pedestri-ans involved in 224 fatal accidents and 5,281

injury accidents. Pedestrians accounted for twothirds of all fatalities. While the number of fatalaccidents involving pedestrians has fluctuatedover recent years, injury accidents declined bymore than 15 percent between 1987 and 1993to make up 35 percent of total injury accidents.

Republic of Korea48 percent fatal accidents

In the Republic of Korea, pedestrians ac-counted for almost half of all road accidentdeaths in 1992 (5,542 of 11,640). This repre-sented a small decrease from the previous yearwhen pedestrian deaths amounted to 6,441 andcomprised more than 52 percent of the total fa-talities (see Lim, 1993).

It should be noted that the proportion ofpedestrians involved in fatal accidents contrastssharply with those of the ESCAP data. Thedata report that only 1 percent of deaths involvedpedestrians in the Republic of Korea, which isan unrealistically low figure, while a paper byPyong Nam Lim at the 1993 Conference onAsian Road Safety (CARS) reported the abovefigures.

Furthermore, more than half of all pedestriancasualties occur among those of working age(21-60) while another 30 percent involvechildren (15 or younger).

Table 3.2: Age Distribution of Pedestrian Casualties

Age Fiji (1992) Malaysia (1994) Papua New Guinea (1992)

Death Injury Death Injury Death Injury(percent) (percent) (percent) (percent) (percent) (percent)

Under 15 16 10 11 10 24 16

16-20 7 14 21 24 9 13

21-60 67 63 58 62 65 70

Over 60 10 3 10 4 1 0

Source: RETA project data.

Table 3.3: Ratio of Pedestrian Fatal Accidents to Pedestrian Injury Accidents

Dhaka, Hong Kong, India Samoa Singapore Sri LankaBangladesh China (1991) (1991) (1993) (1993)

(1993) (1993)

1:0.7 1:24 1:5 1:9 1:15 1:11


9


Singapore25 percent fatal accidents,17 percent injury accidents

In 1993, Singapore reported 241 fatal acci-dents and 5,559 injury accidents, with pedestrians involved in 61 fatal and 930 injury acci-dents. Pedestrian casualties have been decreas-ing in absolute number and relative share oftotal casualties.

Between 1970 and 1986, they dropped from3,250 to 1,293 — i.e., from 32 percent to 15percent of all casualties.

Pedestrian deaths decreased even moresignificantly with only 69 pedestrians killed in1986 (28 percent of the total fatalities)compared to 131 in 1970 (46 percent of the totalfatalities).

In Singapore, adults over the age of 60account for half of all pedestrian deaths.

Taipei,China19 percent fatal accidents,13 percent injury accidents

In 1993, 2,068 road accidents involvingdeaths were reported (along with 628 reportedinjury accidents), equal to a decline of morethan 40 percent since 1989. Pedestrian acci-dents have been reported to be decreasing inrecent years with only 393 fatal pedestrian ac-cidents in 1993 compared to 829 in 1987 andonly 79 pedestrian injury accidents comparedto 914 in 1986.

The unrealistically low pedestrian injuryaccident figure brings the accuracy of the datainto question.

b) Central Asia

PRC11 percent fatal accidents

The 1995 China National Road AccidentStatistics report shows that in 1994, pedes-trians accounted for 11 percent of the PRC’s66,362 road accident deaths, while a Trans-portation Research Record (TRR) article re-por t ed tha t cyc l i s t s and pedes t r i ansaccounted for 60 percent of the country’sdeaths in 1992.

In Beijing, pedestrians were reported as rep-resenting almost 16 percent of road accidentdeaths and nearly 13 percent of injuries between1981 and 1985 (see Navin, 1994).

c) Southeast Asia

Malaysia15 percent fatal accidents,9 percent injury accidents

The 1994 Malaysia Statistical Report onRoad Accidents provided details of the coun-try’s road accidents and related casualties. Outof 5,159 road accident deaths and 43,344 inju-ries, pedestrians accounted for 780 (15 percent)and 4,083 (9 percent), respectively.

Malaysia, with the second highest motorvehicle ownership level (because of the numberof motorcycles) among the developing coun-tries in the Asian and Pacific region, does notappear to have a serious pedestrian accidentproblem. Yet since 1987, pedestrian casualtieshave increased significantly with deaths up by73 percent, and serious injuries by 61 percent,although slight injuries increased by only 10percent. Between 1986 and 1988, pedestriansaccounted for 15 percent of total casualtiescompared to the current rate of 10 percent, im-plying that the accident involvement rate forother road user casualty groups is increasingeven faster than for pedestrians.

One third of all pedestrian casualties inMalaysia involve those under the age of 11,while the peak age is between 6 and10, a groupin which 23 percent of all pedestrian casualtiesoccurred. Total casualties (all modes) peak later(age 16-20) with only 6 percent of total casual-ties involving those under 11 years old.

Thailand9 percent fatal accidents

Almost 85,000 road accidents were reportedto the police in Thailand in 1993, of which 9,496involved a death and 25,330 involved an injury.The 898 pedestrian fatal accidents reportedrepresent more than a doubling in the past threeyears and an eightfold increase since 1988.However, this number represents only 9 percentof the total fatal accidents.

d) South Asia

BangladeshDhaka: 62 percent fatal accidents, 32 percent injury accidents

Accident data are not collected by casualtyclass and so the number of pedestrian casual-

10


ties nationwide is not known. In Dhaka, a re-view of the police’s 1992 First InformationReports concluded that pedestrians accountedfor 60 percent of all deaths. TRL’s Microcom-puter Accident Analysis Package (MAAP) wasintroduced in a pilot project in Dhaka in 1995and has been operational throughout the citysince January 1996. Preliminary findings indi-cate pedestrians accounted for 152 (61 percent)of the 249 deaths, 83 (31 percent) of the 265serious injuries, and 42 (16 percent) of the 262slight injuries. Pedestrians were involved in 143of the 230 fatal accidents (62 percent) and 98 ofthe 306 injury accidents (32 percent) reported.

IndiaUp to 60 percent road deaths

In 1992, 285,900 road accidents were re-ported with 59,720 involving deaths and 276,747involving injuries. Although pedestrian data areunavailable for 1992, 1991 figures show thatpedestrian involvement in accidents amountsto 2 percent of all deaths and injuries. Pedes-trian injury accidents have been reported to bedeclining from 9,014 in 1984 to 5,091 in 1991,while pedestrian fatal accidents have been re-ported to be decreasing from a high in 1985 of1,176 to 982 in 1991. The Road Safety Digestestimated pedestrian deaths to represent be-tween 35 percent and 40 percent of total roaddeaths in metropolitan cities in India and pe-destrian deaths have been found to be as highas 60 percent of India’s total road deaths (RoadSafety Digest, 1995). Hence the extremely lowpercentage of nationally reported pedestrianaccidents must be questioned.

Nepal45 percent fatal accidents,16 percent injury accidents

Since 1995, Nepal’s traffic police have be-gun to collect national details on each injuryroad accident and have found that during Feb-ruary and March 1996, pedestrians accountedfor 49 (45 percent) of 109 deaths and 83 (16percent) of 505 injuries. Pedestrian deathsaccounted for more than a third of all pedestriancasualties.

MAAP was recently introduced inKathmandu and on an adjoining highway. Thefirst accident summary reported pedestrians tobe involved in 44 percent of all accidents and52 percent of deaths in Kathmandu in the lasthalf of 1995. On the Naubise-Mugling section

of the Prithivi Highway, pedestrians wereinvolved in 48 percent of all accidents.

A six-month study into serious bus accidents(i.e., those involving injury) found a total of235 serious bus accidents, 107 of which in-volved pedestrians. Of the 234 deaths causedby bus accidents, pedestrians accounted for61 and 131 of 1,164 injured (11 percent). Thirtypercent of the fatal pedestrian bus accidents oc-curred at night.

PakistanKarachi: 50 percent fatal acidents

No national pedestrian accident data wereavailable for Pakistan, but Karachi, with theonly computerized accident reporting system(MAAP) in the country, was able to providesome pedestrian accident statistics. Pedestriansaccounted for 50 percent of all casualties and50 percent of all deaths, although a high per-centage (47 percent) of pedestrian casualtieswere deaths. In 1991, 84 percent of all pedes-trian casualties were male and 29 percent wereunder the age of 16.

More than 70 percent of pedestrian deathsoccurred while crossing the road and 87 per-cent occurred away from junctions. Only 4 per-cent of pedestrian deaths happened while walk-ing in the road and 28 percent of all pedestriandeaths occurred during the hours of darkness.

Sri Lanka45 percent fatal accidents,48 percent injury accidents

A total of 1,346 fatal road accidents and13,986 injury accidents were reported in 1993 inSri Lanka. Pedestrian fatal accidents havefluctuated in the past five years and 603 werereported in 1993 compared to 627 in 1988,while pedestrian injury accidents reached a highof 6,743 in 1993, an increase of more than 30percent in the last five years. Pedestrian fatalaccidents make up 45 percent of the total fatalroad accidents while pedestrian injuries makeup 48 percent of the total injury road accidents.

e) PDMCs

Fiji43 percent fatal accidents

In 1993, there were 34 pedestrian deaths,154 serious injuries, and 216 slight injuries.

11


The most recent accident statistics for Fiji iden-tified the peak age group for pedestrian casu-alties as between the ages of 6 and 10 (20 per-cent of all pedestrian casualties, yet only 8 per-cent of total casualties). Four out of every fivepedestrian casualties under the age of 11 werehit while crossing the road, with 15 percentoccurring while using a pedestrian crossing.

The younger pedestrian casualties had amuch better survival rate as only 6 percent ofcasualties under the age of 16 died, while 83percent of the pedestrian casualties over the ageof 60 were fatal (Fiji Road Accident Statistics,1994).

Papua New Guinea33 percent fatal accidents,15 percent injury accidents

In 1992, pedestrians accounted for 88deaths, 210 serious injuries, and 142 slight in-juries caused by road accidents. Pedestrianshares of total casualties have remained stableover the past five years, despite pedestriandeaths decreasing by 13 percent to 33 percentof the total deaths, and pedestrian serious inju-ries decreasing by 26 percent to 15 percent ofthe total injuries.

The reductions are similar to the 1987-1992overall decrease in total casualties. Papua NewGuinea pedestrian slight injuries were reportedto be one third less than the pedestrian seriousinjuries and 17 percent of all pedestrian casu-alties were deaths.

One third of all pedestrian casualties in-volved children (15 years or younger) in 1992.Pedestrian casualties peaked in the age group6 to 10, accounting for 19 percent of allpedestrian casualties, and half of all casualtiesoccurred between the ages of 16 and 30.

A recent TRL publication, Pedestrian Acci-dents and Road Safety Education in SelectedDeveloping Countries, presented pedestrianaccident and casualty data from Papua NewGuinea in 1992. Findings included 44 percentof pedestrian deaths occurred during roadcrossings, compared to 36 percent while walkingin the road.

Of the latter, two thirds were reported tohave occurred when the victims walked withtheir backs to traffic. The majority of pedes-trian deaths (82 percent) occurred away fromjunctions, with 59 percent killed in rural areasand 29 percent at night. Males accounted for66 percent of all pedestrian casualties (seeSayer and Downing, 1996).

Samoa50 percent fatal accidents,37 percent injury accidents

In 1993, Samoa had five fatal accidents and45 injury accidents involving pedestrians, ac-counting for half of the 10 total fatalities andmore than one third of the 122 injuries reported.

3.3 Pedestrian Facilities

All countries visited provided a range ofpedestrian facilities and while, in most cases,the overall provision was considered to be in-adequate, there were often good examplesworth considering for use by other countries.The more industrialized countries typicallyhave had a longer history of catering for pe-destrians. Some good and bad examples ofmeasures and facilities are described below.

Pedestrian facilities attempt to balance thecompeting needs of pedestrian safety and con-venience, and vehicle movements. There willalways be a trade-off between these needs, andevaluation is necessary to assess the effective-ness of the various levels of provision. The fol-lowing section illustrates the kinds of measuresthat can and have been implemented to assistpedestrian safety.

However, given the limited transferability ofother countries’ findings and insufficientaccident data, the effectiveness of pedestrianfacilities and remedial measures should beclosely monitored.

Japan, faced with a serious pedestrian safetyproblem in the 1970s, attempted to redress thebalance and undertook intensive constructionof pedestrian facilities.

Between 1971 and 1981, zebra crossingsincreased more than fourfold to 573,000, whilegrade separated pedestrian crossings almostdoubled to 178,000.

Footpaths (and bicycle lanes) were alsogreatly expanded from 26,200 kilometers (km)to 92,700 km. Rather than concentrateexclusively on hazardous locations, Japan tooka proactive approach to improving pedestrianmobility and provided off-street playgroundsfor children, junction channelization, and by-passes (see Koshi, 1984).

This approach has since been recom-mended for developing countries in the Asian andPacific region, as the extent of pedestrianaccident involvement justifies prompt action(see Koshi, 1996).

12


Table 3.4: South African Rural Highway Footpath Design Criteria

Footway Average Pedestrian flow per day daily traffic

Design speed or speed limit Design speed or speed limit(60-80 km/h) (80-120 km/h)

One side 400-1,400 300 200more than 1,400 200 120

Both sides 700-1,400 1,000 600more than 1,400 600 400

Source: CSIR, Pedestrian Facilities Guidelines, 1992.

a) Roadside facilities

(i) Footways on urban roads

As modal segregation offers the best pro-tection for pedestrians, footways are necessaryto provide pedestrians with an alternative tosharing the carriageway with vehicles. Yet de-spite the heavy reliance on walking in develop-ing cities, footpaths often do not receive prior-ity and are constructed only if adequate spaceis available after the necessary road width hasbeen provided. As pedestrian counts are rarelyconducted, the justification for additional orwidened footpaths is rarely documented.

Dhaka, Bangladesh, offers a good exampleof how pedestrian flows are not given adequateattention. Although six out of every ten tripsare by foot (and another two by rickshaw), onlyhalf of the city’s three- and four-lane roads havefootways. A recent United Nations Develop-ment Programme (UNDP) traffic managementstudy found only two out of the six junctionswith the highest traffic flows had footways onall approaches, while 80 percent of therecommended junction designs included foot-way widening or improvements. This impliedthat only 20 percent of junctions studied werethought to have adequate footway facilities (seeDhaka Integrated Transport Study [DITS],1994). In Beijing, PRC, inadequate footways inthe mid-1980s led to carriageway edges beingrailed off for pedestrians.

Begun in the late 1970s, the Calcutta, In-dia, Traffic Engineering Project aimed to giveequal emphasis to footways and carriageways.Surveys of walk speed, density, and flow wereconducted during peak and off-peak hours at20 sites in Calcutta. A minimum design den-sity of 42 pedestrians per minute per meter (ped/min/m) was recommended (equal to a density

of 0.6 pedestrians per square meter [ped/m2])while the desirable density standard was 23ped/min/m (density of 0.3 ped/m2). These val-ues are similar to the density used in Beijing,PRC, of 40 ped/min/m. The City Roads De-sign Principles published by the Beijing CityDesign Committee also recommended aminimum footway width of 3.5 meters (m) with 2m for the footway, 0.8 m for the outer verge,and 0.7 m for the inner verge (Martin Voorheesand Associates [MVA], 1986).

Given the demands on prime urban spacefor economic and social activities in develop-ing countries, footways need also to be main-tained for pedestrian usage if they are to mini-mize conflict with vehicles on the carriageway.While maintaining footways and removingobstructions and encroachment will be difficult,pr ior i ty locat ions such as the busies tjunctions or known accident black spots shouldbe0 kept c lear of obstruct ions so thatfootpaths are available for pedestrians.

Footways should not be allowed to be usedas parking space. Although authorized parkingis rarely provided in developing cities, parkingshould be restricted to roadway space and novehicles, including two-wheelers, should be al-lowed to park on the footway. While footwaysare often required to accommodate street fur-niture, a more recent threat has been from plan-tation boxes whose amenity value is thoughtto justify the intrusion and reduction of footwayspace available for pedestrians. Visibility prob-lems, however, can also result, with bushesmasking pedestrians and restricting pedestri-ans’ vision.

(ii) Footpaths along rural highways

While footways are almost exclusively con-sidered for urban locations in motorized coun-

13


tries, the reliance on walking in developing coun-tries often necessitates siting pedestrianfootways on rural highways. Table 3.4 presentsSouth African design criteria for paved pedes-trian footways outside urban areas.

The South African manual also recommendsrural footways of a minimum width of 1 m and1.5 m in periurban areas. Footways should beseparated from the roadway by at least 3 mwhere possible, and where footways are locatedon sharp bends, guardrails may need to be in-stalled.

Pedestrian needs were not considered in theoriginal planning stage of the Highlands High-way in Papua New Guinea, but as the road isheavily used by pedestrians, the Department ofTransport has since contributed to the fundingof a 13 km footway pilot project. After 5 km offootways had been constructed, two thirds ofpedestrians surveyed chose to use the unsealed1 m wide footway located alongside the drain-age ditch rather than the sealed road. Prelimi-nary analysis of accident data indicated that an-nual pedestrian accidents had decreased by 50percent along that section while increasing atthe control sites without footways by 40 per-cent (see Hills et al., 1991).

In Nepal , a road improvement projectfunded by the Overseas Development Admin-istration (ODA) of the UK has proposed usingthe diversion road ( located at the baseof the embankment) for pedestrian movements. Thediversion road would be made permanent andsealed as a joint 3 m wide footpath and ox cart

path. The total cost for converting 14 km of thediversion road was estimated at aboutUS$70,000.

The provision of footpaths adjacent to shadetrees along rural highways has also been pro-posed in Nepal. To encourage pedestrians awayfrom the road shoulders, a sealed 1.5 m foot-path has been recommended along the high-way near factories. Footpaths must be sealed,otherwise pedestrians will prefer to walk on thecarriageway. Since factories operate 24 hoursa day, an alternative to the road shoulder willbe especially helpful for pedestrian safety atnight.

Highways should also continue footpathsover any bridge encountered (minimum 1.2 mfootway width on bridge). This is particularlyimportant because many bridges in developingcountries are narrower than the approaches andaccidents often occur near or on bridges. Plate3.3 shows a bridge lacking a footway, creatingthe potential for conflict. If the flow ofpedestrians is significant and it is not possibleto mark off an area of the bridge for pedestrianuse, consideration should be given to provid-ing a separate wooden walkway beyond thebridge parapet by cantilevering out from theexisting bridge.

(iii) Shoulders

Where pedestrian traffic is significant butinsufficient to justify a footpath, hard shoul-ders should be sealed in order to provide asmooth compacted surface for pedestrians, asa comparable but preferable alternative to thepaved roadway. Edge lines should also bemarked to provide pedestrians with a visualcue of where the shoulder ends. Coloredshoulder surface treatment is unlikely to beaffordable, but high quality and durable edgeline markings should be marked andmaintained to define the safe areas for pedes-trians.

(iv) Pedestrian precincts

Pedestrianized areas are rarely found in theAsian and Pacific region despite high pedestriandensities and the predominant share of walk trips.Rare examples include the redesign of road spaceand service areas for the convenience and safetyof pedestrians at two rail stations in Mumbai,India, in 1982. In the PRC, some pedestrianizedstreets allow pedal cycle access at certaintimes of the day. Although not an ideal situation,

plate 3.3

Plate 3.3:Lack of footwayprovision on a roadbridge in New Delhi,India.

14


shared cycle use of pedestrianized streets is pref-erable to forcing the NMVs to use dangerous al-ternative routes where they may be in conflictwith motorized vehicles.

Pedestrian precincts improve safety as theytend to be free of accidents. However,pedestrianization holds limited potential for im-proving overall pedestrian safety in the devel-oping world.

b) Crossing facilities

As the majority of pedestrian accidents oc-cur while crossing a road, the need for safe andefficient pedestrian crossing facilities couldarguably be the most important pedestriansafety factor.

(i)Pedestrian refuges

Pedestrian refugesare a common pedes-trian crossing featurein developed coun-tries, yet they arerarely used in the de-veloping countries ofthe Asian and Pacificregion. In SouthAfrica, a properlydesigned refuge islandis assumed to reducepedestrian risk by 50percent (see Centre forScientific and In-dustrial Research[CSIR], 1992).

Where pedestrianand vehicle flows are in-sufficient to justifycontrolled crossings ordualization, pedestrianrefuges allow pedestri-ans to cross the road intwo stages. Pedestriansshould not have tocross more than twolanes of traffic at a time

without a central median provided for refuge.Many of the capital cities in the Asian and Pa-cific region have wide roads, especially Beijing,PRC, and Seoul, Republic of Korea, and wouldbenefit greatly from the introduction of pedes-trian refuges to ease the crossing problems ofpedestrians.

In Bangkok,Thailand, pedestrian crossingnoncompliance was found to be greater on two-way roads than on one-way roads (no relation-ship was found between pedestrian demand andnoncompliance) but this was believed partiallydue to the pedestrian refuges located on two-way roads (pedestrian refuges are not providedon one-way roads). Pedestrian compliance wasfound to be related to the number of traffic lanesto be crossed as well as traffic characteristics(see Jamieson Mackay and Partners [JMP],1988).

Pedestrian refuge islands should, whereverfeasible, be at least 2 m wide and preferably bepainted with reflective markings to maximizeconspicuity. Where space does not allow a 2 mrefuge, a 1.5 m refuge can be used. In Beijing,PRC, the standards are more restrictive with apreferred minimum width of 1.8 m, althoughthe absolute minimum width is recommended tobe no less than 1.2 m.

Pedestrian refuge facilities should be moni-tored to ensure they are achieving the desiredeffect. While the presence of a central refugemay reduce vehicle speeds (by narrowing roadwidth) and simplify the decision making proc-ess (fewer lanes to cross and less approachingtraffic), it subsequently doubles the number ofjudgment calls required and pedestrian safetyimprovement should not automatically be as-sumed.

Central refuges can be designed to allow astaggered crossing, which would guide pedes-trians to face oncoming traffic before they com-plete the second half of the crossing. However,staggered crossings can cause difficulty forpeople walking with a bicycle and are not rec-ommended for joint pedestrian/cycle crossings.

(ii) Medians

The benefits for pedestrians from mediansare rarely appreciated. Where crossing move-ments are uncontrolled (as in the vast majorityof roads in developing cities) and where kerbedmedians are used, the design should be pedes-trian-friendly and avoid introducing unneces-sary obstacles or risks to pedestrians. For ex-ample, in Dhaka, Bangladesh, where narrow andround-topped medians have been used onthe primary roads, they have added to the pe-destrian crossing challenge. Trees have alsobeen planted along some medians, causing vis-ibility problems.

In Beijing, PRC, the few central mediansprovided are usually too narrow at only 1 m to

plate 3.4

Plate 3.4:Pedestrian street inthe PRC. Notevehicles excludedfor certain times ofthe day.

Plate 3.5: Apedestrian crossingpassing over six lanesof traffic in New Delhi,India.

15


1.5 m wide and often are built from barriersthat provide no storage width. Medians con-structed from low-concrete blocks provide littleprotection from traffic. The city’s designprinciples recommended a minimum centralreserve width of 2 m, the absolute minimumallowed at restricted sites is 1.2 m.

The Central Road Research Institute in In-dia is conducting research into the types ofmedians being constructed in New Delhi andtheir relative effectiveness. Negative impactscan arise with dualization as speeds increase.Dualization has long been credited with im-proving road safety by separating traffic flows,preventing head-on collisions, and allowingpedestrians to use the median to cross in twostages. However, in some cases, increasedspeeds may outweigh the safety benefits. Forinstance, a 300 m section of a highway inPapua New Guinea was upgraded to a dualcarriageway and while head-on, rear-end, andsideswipe accidents were reduced significantly,pedestrian accidents and off-road accidentsdoubled. Moreover, while the multivehicle ac-cident types were minor, the sideswipeaccidents, although smaller in number, involved

mainly serious and fatal accidents (see Hills etal., 1991).

(iii) Zebra crossings

Zebra crossings (where pedestrians aresupposed to be granted immediate priority overapproaching vehicles) are provided in some d e-veloping countries, but in the vast majority ofcases, can be quickly dismissed as a tokenmeasure with few, if any, actual benefits. Stud-ies conducted over the years in Dhaka, Bang-ladesh; Surabaya, Indonesia; Karachi, Paki-stan; Colombo, Sri Lanka; and Bangkok,Thailand have confirmed that almost no driv-ers stop for pedestrians using crossings (seeSayer, 1994). While zebra crossings are themost affordable of formalized pedestriancrossings, the disrespect shown to them bypedestrians and drivers renders them of littlepractical use at any cost.

A study conducted more than a decade agoin the Republic of Korea found that almost 40percent of pedestrian accidents occurred on ornear pedestrian crossings (see Ross, 1984). Arecent review of pedestrian facilities in theRepublic of Korea suggested little improvementhad been achieved in pedestrian crossing safety.

While standardized warrants exist for zebracrossings in motorized countries, in most de-veloping countries, zebra crossings are deter-mined on the basis of human judgment. Fewcountries consider pedestrian desire lines; i.e.,the preferred crossing route or the pedestrianand vehicle traffic flows that serve as the basicfactors for warranting zebra crossings in de-veloped countries.

Zebra crossings are not considered appro-priate for heavily traveled roads and manycountries restrict zebra crossings to roads withlow vehicle flows. One of the criteria used inHong Kong, China for installing zebra cross-ings is that the flow should be less than 1,000vehicles per hour. Unfortunately, unprotectedat-grade pedestrian crossings on low volumeroads have been found to be dangerous, asvehicle speeds tend to be high when traffic vol-ume is low. In addition, excess road space andobstructed visibility were found to contributeto pedestrian accident risk.

The zebra crossings with bad accidentrecords in Hong Kong, China had all been in-stalled on wide roads (two-lane roads more than10 m wide or four-lane roads) where on-roadparking restricted visibility (see Craddock,1993).

plate 3.6

plate 3.7

Plate 3.7 (right):Pedestrians having tocross six lanes ofhigh-speed traffic,with no facilities toassist them in NewDelhi, India.

Plate 3.6: A widepedestrian crossingwithout central refugein the Republic ofKorea. Note thevehicle is not givingway to pedestrians.

16


(iv) Traffic signal controlled pedestriancrossings

Unlike zebra crossings, signalized pedes-trian crossings offer the default priority to ve-hicle traffic with pedestrians allowed to crossonly on signals. Unfortunately, like zebra cross-ings, they are not self-enforcing and rely ondriver compliance with crossing regulations.In Bangkok, Thailand, surveys foundsignalized crossings to have a higher compli-ance rate (39 percent to 74 percent) than did

police-controlled (20percent) crossings.

Signalized ped-estrian crossings arerecommended wherepedestrian volumesare high and ped-estrians need to beplatooned to avoidcontinual disruptionof vehicle streams.Pelican crossings(pedestrian-operateds i g n a l - c o n t r o l l e dcrossings) have beenproposed for severallocations in Nepal dueto this factor. Pedes-trian phase length isdetermined by walkspeed and 1.2 metersper second (m/s) is thecommon standard

used, although slightly higher values have beenused in Beijing, PRC (1.3 m/s) and Bangkok,Thailand (1.6 m/s to 1.7 m/s).

Pedestrian accident data analysis of HongKong, China in 1991 determined that a thirdof all pedestrian injuries occurred while pedes-trians were on or near (within 15 m) a signalizedcrossing (see Craddock, 1993). Pedestrianmisuse of the facility was a problem as manypedestrians chose to cross against the pedes-trian signal.

Signalized crossings are costly and requirereliable maintenance support, as any opera-tional problems can quickly cause pedestriansto lose confidence in the facility. Given theprobability of low driver and pedestrian com-pliance and the problems of ensuring adequatemaintenance, signalized crossings are a risk fewdeveloping countries can afford to take.

The effectiveness of zebra and signalized pe-destrian crossings should be evaluated. If pe-

destrians are not using them and drivers notrespecting the crossing, then the facility shouldeither be modified (by adding a guardrail) en-forced (police controlled), or removed. Pedes-trian crossings should be used only where theymake a difference and not where they lowerthe general respect for such measures.

(v) Raised pedestrian crossings

Raised pedestrian crossings were introducedin Karachi, Pakistan, in 1987 at three sites andcompared with two pelican crossings and threecontrol sites. Pedestrians then accounted formore than half of all Karachi’s road accidentdeaths and almost three quarters of pedestriancasualties occurred during road crossings. Sur-veys showed that few vehicles stopped for pe-destrians on the designated crossing and thatwhile drivers were aware of the pedestriancrossing regulations, they chose to violate them.

In addition to forcing drivers to stop for pe-destrians, raised pedestrian crossings had thepotential benefits of:

reducing approach speeds , which re-duced the likelihood of an accident andthe relative injury severity of any acci-dent that did occur. Pedestrian judgmentperception would also benefit as safe gapmisperception increases sharply with ve-hicle speed;

increased pedestrian usage, as thecrossing would be an extension of thefootpath and at the same level, unlikeother crossing points on the road; and

increased visibility and conspicuity, aspedestrian height is raised and so they aremore easily visible to drivers.

The recommended layout of the crossing in-cluded a triangular warning sign proceeded bytwo sets of road humps of increasing heightfollowed by a raised zebra crossing of furtherincreased height. Unfortunately, in implement-ing the measures the preceding alerting andwarning humps were constructed higher thanrecommended while the hump at the crossingwas lower.

Not surprisingly, given the errors in imple-mentation, evaluation of the effectiveness foundmixed results: while accidents were reduced,insufficient accident data from any previousperiod prevented any firm conclusions. Whilevehicle approach speeds were reduced by agreater degree than at the zebra and control

plate 3.8

Plate 3.8: Poorplanning of apedestrian crossing inNew Delhi, India,leading to a heavilyplanted verge.

17


sites, this might have been more due to con-gestion than the humps. Despite the reducedspeeds, the stopping behavior of drivers wasnot improved — less than 1 percent of driverschose to stop for pedestrians using the cross-ing. At the raised crossing that performed theworst, resurfacing removed the markings andalmost certainly contributed to the lack of im-pact on driver behavior.

Although driver compliance remained low,speeds were reduced to below the critical levelof 30 km/h (average speeds at the raised cross-ings were between 20 km/h and 25 km/h) atwhich pedestrians are expected to suffer onlyslight injuries. With a heavy commercial vehi-cle flow, reduced speeds are even more impor-tant for pedestrians.

There is, however, the disadvantage of driv-ers having to reduce speed for the crossings(average delay ranged from a few seconds formotorcycles to 40 seconds for buses at onesite), whether or not pedestrians are using thecrossings. As with all types of uncontrolledcrossings (such as zebra crossings), raised pe-destrian crossings are not appropriate for roadswith heavy or fast vehicle flows, or where pe-destrian flow is heavy and could cause a con-tinual source of disruption to vehicular traffic.

Raised pedestrian crossings have been usedsuccessfully in Fiji on relatively low flow roadsand are common in traffic calming schemes indeveloped countries.

(vi) Grade separated pedestrian crossings

With the increase in traffic levels (both vehi-cle and pedestrian) and lack of compliance withpedestrian crossings, pedestrian overpasses arebeing used more often as they can accommo-date pedestrians without interference fromvehicle flows. The recent Sixth Malaysia Planfunded the construction of 29 pedestrianoverpasses, and permanent and temporary steelfootbridges are being constructed in Dhaka,Bangladesh, and Kathmandu, Nepal.

Overpasses will not be used if they are per-ceived to require longer traveling time. So guard-rails are often required to prevent more con-venient (but more dangerous) at-grade cross-ing. Utilization of pedestrian overpasses inMalaysia ranged from under 3 percent to morethan 96 percent, but no information was givenon the presence of guardrails.

A 6.5 m high pedestrian overpass was con-structed outside a boys’ school in Karachi, Pa-kistan, and to promote usage, 1.5 m guardrails

were put up on the central reservation and for150 m along each side of the road. Despitethese measures, surveys found more than twothirds of pedestrians crossed at grade level byeither climbing over or finding gaps in theguardrails. It was estimated that crossing onthe overpass took 12 seconds longer (total 45seconds) than crossing at grade. The lowvehicle flows were thought to have contributedto the attraction (low-risk perception) and theprevalence of at-grade crossing. No before/afteraccident data was provided (see Sayer andBaguley, 1994).

Pedestrian desire lines (mentioned earlier)need to be considered carefully, as trade-offs oftenwill exist between alternative overpass locations.In Dhaka, Bangladesh, a pedestrian overpass wasconstructed near the airport entrance. While theoverpass was specifically located to access theairport railway station and the designated busstop, pedestrians cross at the nearby roundabout,which has no pedestrian facilities but connectswith the highway. Buses choose to stop alongthe road for such passengers. Other Dhakaoverpasses were constructed finishing on thenearside of the road with pedestrians required tocross busy roads to access the main road or localdevelopments (see DITS, 1994).

Underpasses are less common as they aremore costly and in the subcontinent frequentlysuffer flooding. Dhaka has been constructingthree underpasses, although no publicconsultation was conducted and both utility andflooding problems have been encountered.

(vii) Junctions

According to available data, it appears thatmost pedestrian accidents occur away fromjunctions. While this might be due to dispersedcrossing demand, the lack of pedestrian con-sideration generally shown at junctions givespedestrians little incentive to go out of their wayto cross at a junction.

Crossings should be marked across all ap-proaches (except at T-junctions where theyshould be marked across two approaches) andshould be the width of the footpath. Pedestriancrossing locations must be carefully consideredas they are often sited in awkward places, ob-structed by trees or street furniture. As shownin Plate 3.9, poor guardrail placement can alsoobstruct pedestrian movement.

Beijing, PRC, uses a generous standard of2,700 ped/h/m as the capacity of a pedestriancrossing, and wide pedestrian crossings have

18


been constructed at junctions, which havemoved the vehicle stop line back. As pedestri-ans do not always require the full width, driverviolations often occur and a revised value ofbetween 4,000 ped/h/m and 5,000 ped/h/m hasbeen recommended with crossing widths of 2.5 minstead of 4 m. Pedestrian refuges have alsobeen recommended for junctions to allow a

two-stage crossing,with the refuge sitedfurther back where theroad is narrower andvehicle movementsmore uniform, butwhere pedestrians willstill choose to use therefuge.

At junctions in de-veloped countries, arecent measure hasbeen to extend themain road footpathacross the side roadend (i.e., a raisedcrossover), requiringvehicles to changelevel, however slight,to emphasize ped-estrian priority.

Small radii kerbingat junction cornerswill also help ped-estrians by restrictingturning speeds, al-

though monitoring is required to ensure largevehicles do not mount the kerb and encroachupon pedestrian safe space. In Beijing, a smallcorner radius of 3 m is recommended.

(viii) Traffic signal controlled junctionpedestrian crossings

Pedestrian phases (all red) are extremely rareand pedestrians often must compete with vehi-cle movements, especially those turning near-side. The Calcutta Traffic Engineering Project(CTEP), India, in the late 1970s recommendedbanning left turns where the left turn was heavy;i.e., more than 200 vehicles per hour.

A recent TRL Overseas Road Note, The Useof Traffic Signals in Developing Countries,summarized the range of pedestrian facilitiesat signalized junctions, as shown in Table 3.5.An all red phase built in to the cycle was dis-couraged as it would still be red if no pedestri-ans were crossing and would be likely to incite

driver noncompliance with traffic signals. Aspedestrian movements will tend to be morecritical on one approach rather than beinguniformly distributed, the pedestrian crossingstage should follow the vehicle stage on thecritical approach.

Pedestrians tend to cross according to traf-fic movements and perceived risk rather thanthe displayed signal. This behavior is heavilydependent upon being able to predict the ap-proaching vehicle’s path and Hong Kong,China has found the problems detailed below.

1. Asynchronous vehicle movements at astop line. Pedestrian crossings are oftenlocated immediately in front of the vehi-cle stop line and crossing pedestrians maybe hit by a turning vehicle whose ap-proach is obscured by the lanes of wait-ing traffic. This situation is worsenedwith unsaturated traffic as this encour-ages pedestrians to cross and allows highvehicle approach speeds. One junctionin Hong Kong, China (approach wasonly 20 percent saturated and pedestrianflow was quite low — one tenth that ofbusy shopping areas ) had turning move-ments synchronized with approachmovements and accidents stopped.Accidents were related to method ofcontrol rather than volumes ofpedestrians and vehicles, and asynchronistic method of signal controlwhere vehicle flows are unsaturatedshould be avoided.

2. Road widens into a pedestrian crossing.After turning movements are conducted,the departure side of the intersection al-lows through vehicles to spread out andnot continue straight as expected bypedestrians.

(ix) Roundabouts

Roundabouts hold mixed potential for pe-destrians. While properly designed roundaboutswith adequate deflection will result in reducedvehicle approach speeds, the overall crossingarea is usually increased.

Pedestrian refuges should be incorporatedwith the traffic islands providingchannelization for the vehicles. If no safe andconvenient crossing exists, pedestrians arelikely to choose the most direct route and crosson to the center of the roundabout, a behaviornot unknown in most countries in the region.

plate 3.9

plate 3.10

Plate 3.10:Pedestrian overpassin the PRC withguardrailing tochannel pedestriansover the bridge.

Plate 3.9:Barriers erected toprevent motorvehicles parking onthe footway obstructpedestrian accessto the road.

19


c) General Design Considerations

(i) Visibility

While visibility obstruction cannot fully ex-plain or excuse such extreme low levels of drivercompliance, in Bangkok, Thailand, pedestriandelay was found to be reduced where pedestriancrossings had flashing beacons installed. Indeveloped countries, zigzag markings on theapproach and exit of zebra and signalizedcrossings are used to indicate where motoristsare not allowed to wait, park, or overtake asthey could mask a pedestrian using the crossing.In most developing countries, zebra crossingsdo not have any zigzag markings, additionallighting, or advance warning signs. Road signsare outdated and small in size and almost neverreflectorized.

(ii) Lighting

Visibility is a basic requirement of all pe-destrian crossings and studies in developedcountries have shown that lighting of even amodest standard can reduce nighttime acci-dents on all-purpose main roads by about 30percent. While pedestrian accidents in theregion appear to have a low nighttime involve-ment rate, this could be due to reduced report-ing priorities or opportunities. Furthermore,nighttime accidents may be overrepresented interms of trips but not appreciated due to lackof count data. All grade-level pedestriancrossings (mid block and junction) should bechecked for adequate lighting.

Supplemental lighting may be required toensure the whole crossing is properly lit and thepedestrian does not walk in the shadow.

Table 3.5: Summary of Pedestrian Facilities at Signalized Junctions

Type of facility Characteristic

No pedestrian signal Traffic signals, even without signals for pedestrians, can help pedestrians tocross by creating gaps in traffic streams.Especially applicable where there are refuges and on one-way streets.

Full pedestrian stage All traffic is stopped.Demanded from push buttons.More delay to vehicles than combined vehicle/pedestrian stages.

Parallel pedestrian Combined vehicle/pedestrian stage often accompanied by banning vehiclestage movements.

Useful across one-way streets.

Staggered pedestrian Pedestrians cross one half of the carriageway at a time.facility Large storage area in the center of the carriageway required.

Staggered preferably to face oncoming traffic.

Displaced pedestrian For junctions close to capacity.facility The crossing point is situated away from the junction but within 50 m.

Normal staging arrangements as above.

Source: TRL Overseas Road Note 13, 1996.

plate 3.11 plate 3.12

Plate 3.11: A trafficsignal controlledjunction with apedestrian phase.Note the lack of formalpaved footway leadingto the crossingfacility.

Plate 3.12: Aconfusing pedestriancrossing arrangementat a major junction inNew Delhi, India.

20


which there is usually no scarcity, can be real-located to improve pedestrian safety. Whilemost pedestrian accidents occur at mid-linkcrossings, few traffic police are ever assignedto these locations, although pedestrian accidentsare the most common accident type in most cit-ies in the Asian and Pacific region. In recentyears, road safety has focused on the need toprioritize the more dangerous moving viola-tions, but this has been limited to junctions andhighways and not applied to pedestrian cross-ings.

Traffic police stationed at junctions rarelygive attention to pedestrians and focus on man-aging vehicle flows. Traffic police could in-troduce and enforce turning restrictions socrossing pedestrians have a clear phase on theapproaches.

Traffic police could also prevent buses fromstopping in junctions, and restrict busalightings and departures to designated loca-tions where pedestrian safety is increased.

Accident data should be used to identifytraffic police priorities, which would result inmore attention and priority given to pedestri-ans. Traffic police patrol and point dutyallocation could be used to change staffing fo-cus via patrol times or locations. Traffic regula-tions should be reviewed to assess the deter-rence level of the penalty system regardingpedestrian violations.

ODA is sponsoring a road safety project inNepal under which UK police specialists trainNepali traffic police in organizing enforcementcampaigns, and pedestrian crossings have beenincluded.

b) Education and publicity

Education and publicity programs areneeded for pedestrians and drivers. Traffic edu-cation for children should focus on practicalsurvival skills for everyday situations. Childrenshould be taught to use footpaths and whenfootpaths are not available, they should facetraffic while walking. In Nepal, where foot-paths along rural highways have already beendescribed as scarce, students are being taughtto stay to the left of the edge line when walkingon the road.

Publicity programs are required to teach themasses who are beyond schooling but havenever benefited from any road safety educa-tion. Drivers and pedestrians need to learnabout the vulnerability of pedestrians, how t hey

(iii) Guardrails

Guardrails are used to direct pedestrians to-wards safe crossings and away from high-risklocations. Research conducted three decades agoin the UK found that the presence of guardrailson roads with zebra crossings located every 100yards (92 m) reduced the percentage ofpedestrian crossings between the zebras.

Guardrails should be used at accident blackspots and to restrain pedestrians from crossingon primary or higher class roads where vehi-cle flows should have priority over pedestrians.On local roads, guardrails should be restrictedto accident black spots or vulnerable areas suchas schools or parks. At present, guardrails seemto be used primarily with pedestrianoverpasses. Guardrails will need to be higherthan those used in developed countries as they

should be of adequateheight and design toprevent pedestriansfrom climbing over,and should be see-through so they do nothamper visibility.

3.4IntegratedPrograms

The provision ofVRU facilities does notguarantee effectiveusage and complianceby VRUs and drivers.Education and publicityprograms are needed toimprove understandingand awareness whileenforcement can helpmotivate correct be-havior patterns by per-suasion (verbal warn-ings) and punishment.

To maximize the impact through synergy, engi-neering measures should be coordinated with com-plementary education and enforcement campaigns.Examples are given below.

a) Enforcement

Traffic policing in Asia is characterized bylimited resources and vehicles, but staffing, of

plate 3.13

plate 3.14

Plate 3.13:Visibility ofpedestrians,especially children,has been markedlyimproved for nearsidedrivers in the UK,where special “see-through” fencing suchas Visirail, has beenused

Plate 3.14: Anexample of apedestrian crossing inNew Delhi, India, withcentral guardrailing.Note, however, thenarrow centralmedian.

21


account (most likely) for the most commonroad accident casualties and how drivers andpedestrians can take action to reduce accidentrisks for pedestrians.

In Nepal, under the same ODA projectmentioned above, a pedestrian crossing

plate 3.15

plate 3.16

Plate 3.15:Schoolchildren usinga raised pedestriancrossing in Fiji.

Plate 3.16:Schoolchildrencrossing the roadoutside a schoolunder the direction ofa school crossingpatrol. Note thebarriers at either sideof the crossing.

publicity campaign has been targeted at driversand pedestrians. The publicity campaign wascoordinated with the traffic police’s enforce-ment campaign at pedestrian crossings and theDepartment of Roads’ improvement of, andincrease in, pedestrian crossings in Kathmandu.

22


This chapter reviews safety of NMVs and attempts to identify good practices relatedto provision of facilities for this group of VRUs. Unfortunately, data and research onthis group are far less readily available than for pedestrians, so the depth of analysis

must of necessity be limited to what the project team has been able to gather during theproject. Although not as complete as the review of pedestrian safety, it is intended that thischapter will, nevertheless, provide insights into how the safety problems of NMVs can beaddressed more effectively in the Asian and Pacific region.

4 PEDAL CYCLES AND OTHERNON-MOTORIZED VEHICLES

4.1 Introduction

4.2 NMV Accident and VehicleData

From the limited available data, pedal cy-cles and other NMVs seem to be involved farless in road accidents than other vulnerable road

users (see Tables 2.1 and 2.2). The involve-ment of two- and three-wheeled cycle-basedvehicles in fatal and injury accidents forvarious countries in the Asian and Pacificregion are summarized below in Table 4.1.Accidents involving cycles appear to be high-

Table 4.1: Percentage of Fatal Accidents Involving Cycle-basedVehicles (1) (1993)

Subregion Country Fatal accidents Injury accidents(percent) (percent)

NIEs Hong Kong, China 4 4Republic of Korea 0 0Singapore 8 5Taipei,China 6 5


na(1994)


5(2)

Thailand 4 na


PDMCs Fiji 3 6


na Data not available.(1) Includes two- or three-wheeled cycle-based vehicles.(2) Data relates to casualties at national level.Source: RETA project data.

23


est in Sri Lanka, comprising 17 percent of re-ported fatal accidents and 19 percent reportedinjury accidents.

There are small numbers of cycle-basedvehicles involved in accidents in some of thecountries (notably Fiji, Republic of Korea,Samoa, and Singapore), making meaningful

interpretations difficult. However, it wasobserved that these countries have low numbersof NMVs. Although the PRC’s NationalStatistics show NMVs to be involved in only 9percent of fatal accidents, research from Beijing(see Liu, Shen, and Huang, 1995) shows thatcyclists comprise up to 34 percent of those

Table 4.2: Shares of NMV versus Motor Vehicle Traffic on Roadsin Asian and Pacific Cities

Nonmotorized vehicle Motor vehicle(ex-pedestrian) (percent)

Dhaka (Bangladesh) 51.8 48.2

Phom Penh (Cambodia) 52.1 47.9

Shanghai (People’s 87.2 12.8Republic of China)

Kanpur (India) 55.7 44.3

Surabaya (Indonesia) 15.6 84.4

Tokyo (Japan) 36.1 63.9

Penang (Malaysia) 6.5 93.5

Metro Manila (Philippines) 33.8 66.2

Chiang Mai (Thailand) 2.2 97.8

Hanoi (Viet Nam) 64.3 35.7


Table 4.3: Growth Rates of NMVs in Selected Cities in Recent Years

City Type of NMV Period Average annualgrowth rate

(percent)

Shanghai (People’s Bicycle 1980-1990 14.9Republic of China)

Other NMV 1980-1990 5.3

Kanpur (India) Bicycle 1983-1992 5.3Cycle-rickshaw 1983-1992 2.0Pushcart 1983-1992 1.5

Animal cart 1983-1992 -0.7

Surabaya (Indonesia) Cycle-rickshaw 1985-1992 -0.7

Penang (Malaysia) Cycle-rickshaw 1978-1992 -1.2

Chiang Mai (Thailand) Cycle-rickshaw 1978-1992 8.8


24

City


killed and are involved in as much as 70percent of traffic accidents in the city.

While Table 4.1 shows cycle-based vehicleinvolvement in accidents in various countries,Tables 4.2 and 4.3 permit comparisons of NMVusage and growth in selected cities of theregion. (Table 2.3 earlier showed NMV own-ership levels throughout the region.) Table 4.3shows that the trend in Penang, Malaysia, isthat cycle-rickshaws are declining in numbers,in contrast to the growth in the motorcycle fleetin the country. It is a similar picture in Surabaya,Indonesia. These trends reflect a gradual moveaway from rickshaws towards motorcycle-based vehicles. Following this overview of theAsian and Pacific region as a whole, thedetailed accident and traffic data as availablefor individual countries is presented accordingto the regional subgroupings.

a) NIEs

Hong Kong, ChinaCycles: 4 percent fatal accidents,4 percent injury accidents

In 1993, 336 accidents involving deaths and15,133 accidents involving injuries occurred inHong Kong, China, of which 13 (4 percent)fatal and 556 (4 percent) injury accidents in-volved cycles. No obvious trends are apparentfrom these low figures other than to show thatcycle accidents are not a big problem.

Republic of KoreaCycles: about 0 percent fataland injury accidents

In the Republic of Korea, cycles were in-volved in nine fatal and 268 injury accidentsout of the total 9,471 accidents involving deathsand 251,450 accidents involving injuries in1993. The total number of accidents involvingcycles is insignificant, although fatal accidentsinvolving cycles actually decreased by 74 per-cent from 1991 to 1993 and injury accidentsdecreased by 67 percent over the same period.

SingaporeCycles: 8 percent fatal accidents,5 percent injury accidents

In 1993, 19 fatal accidents and 265 injuryaccidents involving cycles occurred in Singa-

pore out of a total of 241 and 5,559 accidentsinvolving deaths and injury, respectively. Inpercentage terms, cyclists were involved in 8percent of the total fatal accidents and 5 per-cent of the total injury accidents. Since 1989,the number of fatal accidents involving cyclistshas risen, along with a general increase in theirrelative share of the fatal accidents.

Taipei,ChinaCycles: 6 percent fatal accidents,5 percent injury accidents

In 1993, 128 fatal accidents involving cy-clists were reported out of the total 2,068 fatalaccidents, representing 6 percent of the total.The number of fatal accidents has decreased ata fairly constant rate since 1990. Only 31 in-jury accidents (5 percent) involving cycles wererecorded in 1993 out of the 628 total injuryaccidents recorded. The low injury accident fig-ures call into question the accuracy of the data.

b) Central Asia

PRCNMVs: 9 percent deaths(Beijing cyclists: 34 percent deaths)

The 1995 Chinese National Road AccidentStatistics shows that in 1994, 5,675 road deathsinvolved NMVs, making up only 9 percent ofthe total road fatalities. However, a TRR arti-cle by Liu, Shen, and Huang in 1995 relatingto Beijing reports that cyclists were involvedin 34 percent of accident deaths (1992); fur-thermore, between 1981 and 1990, about 70percent of the total traffic accidents involvedcycles.

Analysis of fatal accidents in Beijing re-vealed that between 1981 and 1992, the aver-age rate of cycle involvement in fatal accidentswas more than 30 percent, with a maximuminvolvement rate in 1981 of 43 percent and aminimum level of 25 percent in 1989. Despitea decreasing trend in cycle involvement in fa-tal accidents up to 1989, cyclist involvementin fatal accidents rose to 34 percent in 1992.

Analysis of age-related data for fatal acci-dents involving cyclists in Beijing revealed thatriders aged over 60 made up only 11 percent ofthe population and contributed to only 5 percentof the trips made. But this group was involvedin more than a quarter of the fatal accidents

25


involving bicycles — clearly a target groupneeding special attention.

More than 80 percent of fatal accidents in-volving cycles occurred on arterial andsubarterial roads in Beijing. This high severityrate was due to the higher operating motor ve-hicle speeds and a large number of junctionsalong their lengths.

Indeed, 54 percent of these fatal accidentsoccurred at intersections, despite many of theintersections being signalized.

Table 4.4 shows the major causes of fatalaccidents involving cycles in Beijing between1981 and 1990. The majority of fatal accidentsinvolved cyclists not yielding to motorized ve-hicles or carrying out turns, especially acrossoncoming vehicles. A major problem that has

been identified is the mixing of motor vehicleand cycle traffic, especially at junctions wheredifferences in speeds clearly create problems.

The article also notes that farmers had thehighest involvement rate in fatal bicycle acci-dents, due to their poor knowledge of trafficlaws, the poor and unsafe condition of theirbicycles, higher motor vehicle speeds in ruralareas, and the difficulty in receiving promptmedical treatment in rural areas. As a result,rural areas have a higher percentage of fatalaccidents involving cycles than urban areas, de-spite only a third of cycles being owned in ru-ral areas.

One of the main causes of cycles being in-volved in a high number of accidents in Beijingis the quantity of cycles. Liu, Shen, and Huang

Table 4.5: Registered Vehicles by Type and Yearin the PRC

Year Motor vehicle Bicycle Total Bicycles as(10,000) (10,000) (10,000) percent of total

1985 655.65 22,364 23,019.65 97.2

1986 819.07 25,803 26,622.07 96.9

1987 1,061.03 29,313 30,374.03 96.5

1988 1,190.20 33,312 34,502.20 96.6

1989 1,318.53 36,515 37,833.53 96.5

1990 1,476.26 39,099 40,575.26 96.4

1991 1,657.66 41,979 43,636.66 96.2

1992 1,945.03 45,076 47,021.03 95.9

Source: National Road Accident Statistics, 1995.

26

Table 4.4: Causes of Accidents Involving Bicycle Deathsin Beijing, PRC, (1981-1990)

Cause of Accident Involvement (percent)

Unyielding to other traffic 27.0

Sudden turning 26.3Riding on motorway 16.1Crossing more than four lanes 8.5

Riding the wrong way 7.7Reckless riding 5.3

Riding with a passenger 1.9Poor riding skills 1.2Riding under the influence 0.9

Others 2.3

Source: TRR No. 1487, article by Liu, Shen, and Huang, 1995.


claim that as the number of daily cycle tripsincreases, so do the chances of motor vehicleand cycle collisions.

In 1992, Beijing had nearly 8 million regis-tered bicycles and 700,000 registered motorvehicles.

Table 4.5 shows the number of registeredvehicles by type and year throughout the wholeof the PRC. It shows that motorized vehiclesand bicycle numbers have increased between1985 and 1992, the rate of growth being aboutthe same for both types of vehicle.

Although nonmotorized vehicles have in-creased in numbers, their proportion on theNational Major Highways of the PRC (NMHC)decreased between 1988 and 1992 (see Liu,Shen, and Huang, 1995). Table 4.6 shows theshare of NMV traffic on NMHCs. It ispredicted that this proportional decrease willcontinue into the year 2000.

It should be noted, however, that the PRC’sprovinces each have different proportions ofNMV traffic on the NMHCs.

c) Southeast Asia

IndonesiaCyclists: 9 percent fatal accidents,

7 percent injury accidents

Thompson and Rudjito in Traffic AccidentInvestigation in Indonesia (1991) report thatcyclists comprise 7 percent of injuries to roadusers and 9 percent of all deaths.

National data on the number and size of thenational NMV fleet in Indonesia was unavail-able. However, limited data with regard toSurabaya from a PADECO Co., Ltd. (1995) re-port revealed that NMVs made up only 16 per-cent of traffic on roads despite 45 percent ofthe vehicle fleet belonging to the NMV cat-egory. The majority of the NMV fleet comprisesbicycles, which make up 40 percent of vehi-cles owned, and rickshaws, at less than 5percent. No clear national policy exists withrespect to NMV provision except to recognizethe transport requirements of individual cities.

Table 4.7: Cyclist (NMV) Fatal Accidents by Year in Thailand

Year Fatal accidents involving Total fatal accidents Percent of fatal accidentscycles involving cycles

1988 489 2,015 24

1989 177 6,963 3

1990 129 5,765 2

1991 104 6,319 2

1992 285 8,184 3

1993 338 9,496 4


Table 4.6: Percentage of Nonmotorized Vehicles on NMHC Highways

Year Bicycle traffic Fatal accident Total NMV traffic(person-drawn and animal-drawn

vehicle traffic)

1988 8.9 6.8 15.7

1989 8.4 6.5 14.9

1990 8.2 6.1 14.3

1991 7.7 6.0 13.8

1992 6.8 6.4 13.2

Source: Liu, Shen, and Huang, 1995.

27


Hence, Indonesian cities were observed toexhibit marked differences in the numbers ofNMVs and the facilities provided toaccommodate them. In Jakarta, for example,cycle-rickshaws have been banned and re-placed by motorized motorcycle taxis, resultingin a highly motorized and polluted trafficenvironment that is unattractive to other two-wheeled and nonmotorized vehicles. However,

some NMVs, such as handcarts and cycle-rickshaws, do still appear to operate within Jakarta.

MalaysiaBicycles: 6 percent fatal accidents, 5 percentinjury accidents

In 1994, 326 (6 percent) road accident deathsand 2,177 (5 percent) injuries involving bicy-

Table 4.8: Vehicles Registered by Type and Year in Thailand

Year Registered NMVs Total registered NMVs as percentage ofvehicles the total vehicles

1990 50,270 07,592,100 0.7

1991 53,930 08,481,000 2.0

1992 37,300 09,595,200 3.0

1993 39,300 11,102,000 4.0

Source: RETA Project data.

Table 4.9: Vehicles Responsible for, or Involved in, Traffic Accidents in Hanoi,Viet Nam (1991)

Vehicle Number Percent

Bicycle and cyclo 48 11.0

Motorcycle 265 60.6

Xelam 21 0 4.8

Automobile 13 0 3.0

Truck and bus 90 20.6

TOTAL 437 100.0

Source: TRR No. 1441 (1994) .

Table 4.10: Vehicles by Type in Hanoi, Viet Nam (June 1991)

Vehicle Type Number Percent

Bicycle 1,000,000 84.6

Cyclo 5,100 0.4

Oxcart 0 -

Motorcycle 128,000 10.8

Xelam 1,667 0.1

Auto Rickshaw 0 -

Automobile 35,000 3.9

Truck 11,800 1.0

Bus 160 0.0

TOTAL 1,180,860 100.0

Source: TRR No. 1441 (1994).

28


cles were reported out of a total 5,159 and43,344 road deaths and injuries, respectively,in Malaysia. No obvious trend exists for eithercasualty figures involving bicycles.

ThailandCyclists and NMV three-wheelers:4 percent of fatal accidents

Only a small amount of data has been col-lected for Thailand, indicating that 4 percentof accidents involve cyclists or nonmotorizedthree-wheelers. Despite this, a big increase inthe number of fatal accidents involving bicy-cles occurred in 1993, compared to 1991 (Table4.7). The 1988 figure is also abnormally high,which raises questions about data accuracy andconsistency.

NMVs comprise only a tiny proportion ofregistered vehicles in Thailand, as indicated inTable 4.8.

Viet NamNMVs: 11 percent of all accidents

NMVs are commonly perceived by officialsin Viet Nam as being responsible for “50 to 90percent of all traffic accidents” (PADECO Co.,Ltd.). Statistics gathered by the Hanoi TrafficPolice and published in TRR No. 1441, how-ever, indicate that they were involved in only11 percent of all reported accidents.Motorcycles were, in fact, responsible for morethan 60 percent of the accidents (see Table 4.9).

In Viet Nam, particularly in Hanoi, residents

have traditionally depended upon NMVs, es-pecially bicycles and cycle-rickshaws (cyclos).However, rapid motorization is occurring as thecountry develops and the main growth is in theuse of motorcycles. Despite this, traffic countsshow that NMVs still account for up to 70 per-cent of vehicle volumes. Vehicle types and us-age in Hanoi are shown in Tables 4.10 and 4.11.

In Hanoi, “plans have been made to abolishall NMVs from operating within the city by theyear 2004” (see Bell and Kuranami, TRR No.1441, 1994). However, some provision forNMVs does exist, for example, physicallysegregated lanes for motorized and slow-mov-ing vehicles. Furthermore, trucks and buseshave been banned from using certain roads (seeBell and Kuranami, TRR No. 1441 , 1994).Consequently, although official governmentpolicy appears to be aimed at “abolishing”NMVs, the reality is that they still form a majorpart of the traffic stream in much of Viet Nam.

d) South Asia

BangladeshDhaka: NMVs,13.5 percent accidents

Over nine months in 1995 to 1996, a pilotaccident reporting project in Dhaka revealed260 registered accidents; of these, NMVs ac-counted for 35 accidents (13.5 percent). Morethan 50 percent of the NMV accidents were fa-tal, 80 percent of these involved a rickshaw, 43

Table 4.11: Pedestrian and Vehicle Volumes in Hanoi (Based on Peak-hour Surveys at TwoLocations in June 1992)

Mode Number Percent

Pedestrian 582 7.1

Handcart 0 -

Bicycle 4,578 55.8

Cyclo 321 3.9

Motorcycle 2,377 29.0

Xelam 49 0.6

Auto rickshaw 0 -

Automobile and truck 230 2.8

Bus 66 0.8

TOTAL 8,203 100

- = magnitude zeroSource: TRR No. 1441 (1994) .

29


percent of the rickshaw accidents occurred atT-junctions, and more than half involved a rear-end collision. The data were produced by theMAAP recently installed in Dhaka.

Accident statistics, with respect to cycle-rickshaws, are scarce and unreliable(underreporting is a big problem), but R.Gallagher, in The Cycle-rickshaws of Bangla-desh (1992), estimates that in 1986 and 1987,cycle-rickshaws accounted for about 10 percentof road deaths in Dhaka (based on newspaperreports). Bus, trucks, and minibuses accountedfor 97 percent of rickshaw-related deaths and100 percent of bicycle deaths.

Gallagher further estimates that there areabout 700,000 cycle-rickshaws in the country,with between 150,000 and 200,000 in Dhakaalone — more than half of all Dhaka’s vehicles.Bicycle numbers in Dhaka are between 33,000and 50,000. (Registered cycle-rickshaws num-bered 100,000.) However, the observed usagewas low. But provincial towns were stated tohave a much higher percentage of bicycles inuse. The number of cycle-rickshaws is rising

(see Table 4.12) and it is estimated that by theyear 2000, they will total more than 1.25 millionthroughout the country.

The 1993 Statistics Yearbook gives the fol-lowing table of total registered cycle-rickshawsin 19 selected municipal corporations, includ-ing Dhaka.

Manual traffic counts made on several ma-jor national roads (1995) show that on aver-age, NMVs made up 49 percent of the totaltraffic. NMVs were split between cyclists (29percent), cycle-rickshaws (66 percent), andcarts (5 percent).

It should be noted that there is a generalpolicy in Bangladesh of rickshaw reduction. “InApril 1987, the government of Dacca, Bangla-desh, announced plans to completely banpedicabs from the city on safety grounds, al-though they employed more than 100,000 peo-ple” (TRR No. 1294, 1991). This policy is stillongoing, as Plate 4.1 shows, but it is a highlypolitical problem. NMV measures observed inDhaka include both the banning of NMVs oncertain sections of road, and the provision ofsegregated facilities, mostly by carriagewaymarkings.

Numerous heavily-laden pushcarts were alsoobserved throughout the city, but especially inthe old part of Dhaka where carriageway widthswere extremely narrow in sections, due in partto traders encroaching on to the highway. Con-gestion was also particularly severe and exac-erbated by the horse carts.

Sri LankaPedal cycles: 17 percent fatal accidents,7 percent all accidents

A report by the Sri Lanka Traffic Police andthe Ministry of Transport revealed that between1986 and 1990, 15 percent of deaths were pedalcyclists and that cycles were involved in 6 per-cent of all accidents. Cyclists were involved in

plate 4.1

Plate 4.1:Confiscatedrickshaws in Dhaka,Bangladesh.

Table 4.12: Registered Vehicles in 19 Selected Municipal Corporations (Bangladesh)

Year Cycle-rickshaws Motor vehicles Cycle-rickshaws aspercent of total

1988 191,306 270,860 41

1989 270,430 315,530 46

1990 271,858 341,648 44

1991 295,334 364,626 45


30


19 percent of reported injury accidents and 17percent of reported fatal accidents. This high-lights the high risk of death for cyclists involvedin accidents.

Tables 4.13 and 4.14 show pedal cycle in-volvement in accidents and fatalities.

From the tables above it can be seen that thenumber of cyclists involved in accidentsreached a peak in 1990 and then fell to a low in1991.

Fatal accidents involving bicycles have de-creased since 1990 while their percentage ofinvolvement in fatal accidents has also de-creased since 1991.

e) PDMCsMinimal bicycle accidents

Of the 11 island nations representing thePDMCs, data were available only for Fiji. In1993, only two fatal accidents involvingbicycles were recorded. This followed a lownumber of fatal accidents involving bicyclesover the previous five years.

4.3 NMV Facilities

As illustrated in the previous section, largenumbers of NMVs are in daily use in manyAsian countries. Despite their obvious impor-tance as a transport mode, few cities in theregion specifically provide measures to assistNMVs and many roads and networks withinsome Asian cities appear to have been designedwithout any consideration for their needs. In-deed, not one city was observed to provide anetwork of linked measures for NMV users toassist popular journey patterns. However, someinteresting and useful solutions to help deal withthe high NMV levels do exist and can be ob-served to be operating successfully throughoutthe region.

The following section is intended to illus-trate and highlight successful NMV facilitiesfrom around the region with respect to theirgeneral design requirements. However, this re-port is not intended to be a design manual andseeks only to indicate the general form thatthese measures can take. The facilities dis-cussed are suitable for pedal cyclists and other

Table 4.13: Pedal Cyclists as Percentage of all Reported Accidents in Sri Lanka

Year Cycle accidents Total accidents Cycle accidents aspercentage of total

accidents

1988 2,191 33,028 7

1989 1,960 31,696 6

1990 2,366 34,463 7

1991 1,842 34,144 5

1992 2,574 37,777 7

1993 2,900 41,495 7

Source: Sri Lanka Traffic Police, 1996.

Table 4.14: Fatal Accidents Involving Pedal Cycles in Sri Lanka

Year Cycle fatalities Total fatalities Cycle fatalities aspercentage of total

fatalities

1988 226 1,365 17

1989 224 1,454 15

1990 265 1,714 16

1991 241 1,255 19

1992 231 1,302 18

1993 222 1,346 17

Source: Sri Lankan Traffic Police and the Ministry of Transport.

31


NMVs, but it should be noted that the designof any facility should take into account the ex-pected NMV traffic composition.

While engineering measures assist with safertravel for NMV users, complementary educa-tion and enforcement programs increase theireffectiveness. This section focuses largely onengineering measures, but some guidance isalso provided on related education and enforce-ment issues.

a) Road links

Various facilities can be provided along roadlinks (sections of a road between intersections)to assist NMVs. These facilities can be catego-rized into segregated measures, either by physi-cal or visual means, or those measures that helpNMVs within a mixed traffic environment.

Segregational measures usually separate mo-torized and nonmotorized traffic, but at timesthey divide four-wheeled motorized vehiclesfrom the rest of the traffic. The type of

segregation most suitable is dependent on thefactors associated with the road hierarchy andland use, such as motor vehicle traffic speedsand volumes, traffic composition, and vehicleaccesses. High cycle flows may warrant provi-sion of a segregated facility and will help gov-ern the actual physical dimensions of the facil-ity provided. Total segregation (where NMVsrun on a completely separate network of pathsto motor vehicles) can also be provided, but noexamples of this type have been identified inthe region.

Where NMV traffic has been banned fromusing certain roads, it is important that an alter-native network exist for NMV users and thatthese alternative routes are attractive in termsof safety, comfort, and journey time.

(i) Physical segregation

Physical segregation of nonmotorized andmotorized vehicles using a barrier or raised trackis intended to maximize cyclist safety along roadlinks that have high vehicle speeds, high vehicleflows, or a corresponding mixture of both. Theactual barrier or height difference between therunning surfaces prevents the interaction of thetwo types of vehicle. The infrastructure in-volved in physical segregation requires highinitial implementation costs in comparison tovisual segregation.

Segregation through raised kerbs helps de-lineate dedicated cycle lanes alongside majorroads. Safety can be increased with the provi-sion of adjacent verges in order to further sepa-rate the two modes of vehicle (see Plates 4.2and 4.3). A raised track or verge edge heightshould be sufficient to discourage motor vehi-cle encroachment onto the segregated track.Furthermore, it is also often necessary to pre-vent illegal use of these paths by motor vehi-cles by placing bollards at either end of the seg-regated section. Limited examples of raised andkerb separated lanes were observed in Dhaka,

plate 4.2 plate 4.3

plate 4.4

Plate 4.2:Physical segregationby a tree-lined vergein Ho Chi Minh City,Viet Nam.

Plate 4.3:Continuous physicalsegregation bykerbing along anarterial road inYogyakarta,Indonesia.

Plate 4.4:Railings used tosegregate cyclists andmotorized vehicles inthe PRC.

32


plate 4.6

Bangladesh, such as the raised pavements atthe High Court and the segregated lane onMirpur Street; on Malioboro Street in Yog-yakarta, Indonesia; and in Ho Chi Minh City,Viet Nam.

Alternative segregational methods can in-clude periodic barriers, spaced at intervals along

plate 4.5

the route that may be in the form of upstandingblock work, bollards, posts, or even old oildrums, as well as actual continuous barrierssuch as low railings and walls (see Plate 4.4).In the PRC and Viet Nam, effective continuouslane barriers have been erected using concreteblocks joined by metal poles that can be dis-mantled and are moveable when required andyet offer an effective near permanent barrier atmany locations in cities in the PRC (see Plate5.6).

Physically segregated paths can be used fortwo-way cycle flows, but their design must en-sure that a satisfactory width has been providedin order to accommodate two- and three-wheeled NMVs, along with any overtakingmaneuvers (see Plate 4.5). Particular care alsoneeds to be taken in the design of merging anddiverging areas where nonmotorized and mo-torized vehicles mix.

Segregated facilities for NMVs at pinchpoints, such as road bridges throughout theregion, are rare. However, a few good exam-ples do exist as in Dhaka, Bangladesh (seePlate 4.6).

Yogyakarta, Indonesia, provides good exam-ples of slow moving vehicle (SMV) carriage-way separators, which are intended for flows intwo-way and one-way directions according tothe road hierarchy. These lanes allow safe anddirect travel for NMVs in the town center and onthe arterial routes into and out of the city. Onlanes with access for service vehicles, protectedloading bays are provided in order to preventdisruption of the SMV flows. Lane widthstended to be about 5 m wide in Yogyakarta,allowing three cycle-rickshaws to pass side byside. Signs limited the lanes’ use to NMVs andmotorcycles, with cars and trucks explicitlybanned. The 1995 Yogyakarta TransportationMaster Plan also recommends the provision offurther SMV lanes (see Figure 4.1).

The proposed 1994 geometric design stand-ards for Bangladesh (Road Materials and

Plate 4.5:Slow-moving vehiclelane for two-waytraffic in Yogyakarta,Indonesia.

Plate 4.6 (far right):Physical segregationfor NMVs over a roadbridge in Dhaka,Bangladesh.

Figure 4.1:Proposed cross-sections inYogyakarta,Indonesia.

Source: Yogyakarta Urban Development Project, 1995.

33


Standards Study) include cross-section recom-mendations for the inclusion of NMV lanes onnational, regional, and feeder roads. Thismanual not only provides design dimensionsbut also capacities for the road categories un-der consideration. The provision of cart or bulltracks is also catered for on feeder roads (seeFigure 4.2c).

(ii) Visual segregation

Visual segregation of nonmotorized and mo-torized vehicle traffic is provided by white linemarkings on the carriageway. These nearsidecarriageway markings can take the form of ei-ther broken white lines, termed advisory cyclelanes (see Plate 4.7) in the developed countries,or continuous solid white lines, termed man-

plate 4.8plate 4.7

datory cycle lanes (see Plate 4.8). Cyclists arerecommended to use either of these lanes wherethey exist; motor vehicles are not permitted toenter the mandatory lane and may enter theadvisory lane only when necessary, and only ifcyclists are not impeded. Such facilities aredependent on driver compliance and effectiveenforcement of the traffic regulations.

The type of lane to be implemented is de-pendent on road hierarchy and land use factorssuch as carriageway width and parking require-ments in the area. For instance, parallel park-ing alongside the pavement will discouragecycle use of the lane and render it ineffective.Flows along these lanes should be in the gen-eral traffic direction only. So lane widths shouldnot be so wide as to encourage two-way cycleflows or to encourage illegal use by motor vehi-cles during times of congested traffic.

Markings, and hence the integrity of the ac-tual lane can be enhanced by providing widemarkings, double-line markings or evenhatched areas to denote the separation. Alter-natively, yellow paint has also been used to helpdenote the lane, as in Metro Manila,Philippines (PADECO Co., Ltd.). There arelimited examples of white lines throughout theregion, such as in Bangladesh and Indonesia.Lanes observed in Dhaka, Bangladesh, tendedto be 1.5 m to 2 m wide, making them suitablefor bicycles, but wide enough for only a singlerickshaw, which forces overtaking NMVs tostray outside the lane.

Contra-flow cycle lanes have been imple-mented in cities in the PRC, such as Shanghai,and generally appear to work well with onlyvisual segregation to separate the traffic flows(see Plate 4.9). However, the use of suchmeasures and implementation experience inAsian developing countries outside of the PRCis not known. A system of one-way motorizedtraffic and one-way NMV traffic is also in usein Shanghai, in addition to the traditional con-tra-flow system, (see Plate 4.10).

Figure 4.2:Proposed cross-sections inBangladesh.

Source: Road Materials and Standards Study, Bangladesh, 1994.

Plate 4.7:Visual segregation forSMVs using brokenwhite lines inYogyakarta,Indonesia.

Plate 4.8 (far right):Visual segregation forSMVs using a solidwhite line in Dhaka,Bangladesh.

34


plate 4.10plate 4.9

Contra-flow cycle lanes are usually denotedby a continuous white line, although standardsin developed countries recommend that physi-cal segregation be provided at the start and endof such sections. Contra-flow cycle lanes canhelp provide an area-wide network for NMVusers by creating direct and cohesive routes.Safe design requires crucial consideration ofthe lane widths so as to reduce the attraction oftraveling out of the lanes into the main carriage-way. PADECO Co., Ltd. proposes that with-flow lanes should also be provided in the op-posing direction to the contra-flow lane to in-crease safety; however, this would be at thefurther expense of the effective motor vehiclecarriageway width.

(iii) Mixed traffic

The planned use of mixed traffic ofnonmotorized and motorized vehicles can berelatively safe under certain conditions, suchas low motor vehicle speeds and flows, alongwith an appropriate traffic stream composition.Safety aspects relating to VRUs in terms ofspeed and traffic reduction measures are dis-cussed elsewhere within this report, along withways of reducing the interaction between in-compatible vehicles.

An increasing problem is the growingnumber of accidents occurring on rural roadshaving their surface condition improved by roadrehabilitation projects. This encourages highervehicle speeds, which accentuate even furtherthe differences between motor vehicles andNMVs. Additional safety features need to beincorporated when such rehabilitation projectsare undertaken to avoid road safety problems.This is particularly so where large numbers ofpedestrians, NMVs, or SMVs can be expected.In Bangladesh, for example, the mixed envi-ronment for cyclists was enhanced by the oc-casional provision of speed humps to slowdown motorized traffic. This is of particular

value when rural roads pass through townshipsor communities.

The provision of facilities for joint use bypedestrians and NMVs without segregation isnot recommended where high pedestrian flowsexist and should not be the preferred solutionwhere alternatives exist. This approach was,however, observed on busy radial routes inShanghai, PRC.

Where mixed traffic does exist, the widen-ing of nearside road lanes by altering the posi-tion of line markings is being carried out inAustralia, Europe, and the US. This practiceallows a greater clearance distance betweencyclists and overtaking motor vehicles. Caremust be taken, however, to ensure this insidelane width is of a size to encourage use by tworows of motor vehicles during periods of trafficcongestion.

b) Junctions and crossings

The potential for conflicts betweennonmotorized and motor vehicles is greatest atintersections where segregation of traffic is notalways possible or suitable. As a result, con-flicts at junctions account for a large propor-tion of all cycle accidents, as shown in researchcarried out in the PRC (see Table 4.4). Plates4.11 and 4.13 show examples of pedal cycleand motor vehicle conflicts at intersections.However, as part of the network design process,NMVs must be given full and properconsideration at junctions, taking into accounttheir specific vehicle characteristics. Plate 4.12shows an example of how poor planning in thePRC has caused a potential conflict point dueto the right-turning traffic having to cut acrossthe cycle lane with no priority markings or signsprovided.

Measures to assist cyclists and other NMVsat priority junctions, signal controlled junc-tions, and roundabouts are considered below,

Plate 4.9:Contra-flow cyclelanes in Shanghai,PRC.

Plate 4.10 (far right):One-way motorvehicle flow and one-way cycle flow systemin Shanghai, PRC.

35


plate 4.11

along with grade-separated facilities such asfootbridges and underpasses.

All intersections should be appropriatelymarked and signposted in order to provide clearinstructions and directions to motor vehicledrivers and NMV users alike. The design of anyjunction with specific NMV provision shouldinclude all the normal geometric factors, includ-ing sightlines and visibility requirements thatwould be considered for any other vehicle.

(i) Priority junctions

NMV users should expect the same privi-leges as drivers of motorized vehicles andwhenever possible, should not be preventedfrom carrying out a full range of legal turningmaneuvers at junctions. In order to carry thesemaneuvers, weaving by cyclists and motor ve-hicle drivers (in order to be in an appropriateturning lane) is common practice so junctionsneed careful design. For example, turningNMVs are often extremely exposed when wait-ing in the center of the carriageway for a gap

plate 4.13

in the traffic. In order to reduce conflicts, it issuggested that priority junctions be designedin order to minimize the interaction time by re-ducing the distances to be crossed, for exam-ple by reducing turning radii. Short turningmovements can be helped by leading cyclelanes away from the junction and on to the ad-jacent junction arm. No specific operating ex-amples of measures to assist NMVs at priorityjunctions in the region have been identified todate during the study.

Turning restrictions via physical measureson motor vehicles can have beneficial effectson safety as the interaction between vehicles isreduced. Reduced vehicle speeds and reduc-tions in motor vehicle priority also have ben-eficial safety consequences for NMVs (see sec-tion on Traffic Calming in Chapter 6). Trafficpolice were commonly observed controllingflows and access at junctions in Bangladesh,PRC, and other countries. When the police areproperly trained, this can be an effective methodof traffic control with particular benefits toNMVs.

(ii) Signal junctions

In Beijing, PRC, some bicycle traffic signalsexist on roads with heavy cycle flows (see Plate4.14). These signals are cycle-traffic actuateddue to the use of bicycle sensitive inductiveloops in the road surface (PADECO Co., Ltd.,1993). It should be noted, however, that complexsignal-controlled junctions are a costlysolution. Given the widespread NMV usagethroughout Asia, other less expensive trafficmanagement alternatives should first be

plate 4.12

Plate 4.11: A potentialconflict at a junctionbetween a cyclist andtruck in Hanoi, VietNam.

Plate 4.13:Conflicts betweencyclists,motorcyclists, andmotor vehicles at ajunction in Shanghai,PRC

Plate 4.12:Example of poorplanning at a junctionin Shanghai, PRC.Note that trafficturning right must cutacross the adjacentcycle lane.

36


considered at signal-controlled junctions wheretraffic levels are high. Examples are givenbelow.

The banning of all opposing traffic move-ments, regardless of vehicle type, helps reducethe number of potential conflicts. This has beencarried out successfully in Ho Chi Minh City,Viet Nam, where opposing turning maneuvershave been banned over a series of junctionsbefore a roundabout, for example on Le Lo’iRoad and Ham High Road. However, cyclistsare particularly vulnerable at roundabouts andit is suggested that variations on the abovemethod be considered. These variations main-tain the banned opposing vehicle turning move-ment but allow NMVs to cross the major roadfrom the minor road using a straight crossingmaneuver. This process involves taking theNMV off the major road either before (G turn)or after (Q turn) the junction and then leadingthe NMVs to the minor road arm of the junc-tion via a series of turns with the general trafficflow. Figures 4.3 and 4.4 illustrate the situationwhere traffic drives on the right. Note that theleft turn is prohibited at point A and traffic isrerouted in order to allow that maneuver.

Conflict reductions have also been achievedin the UK and other developed countries byaltering the layouts of some junctions in orderto allow cyclists access to a waiting area, or

reservoir, at the front of the traffic queue butbehind a separate cycle stop line (see Plate4.15). This advance stop line provides thechance for cyclists to carry out their maneuverat the head of the traffic with increased safetyand negligible effect on motorized traffic. Natu-rally, enforcement and driver awareness ofthese areas to prevent motor vehicle access isnecessary and the installation of colored sur-face materials or clear markings can be useful.Similarly, enforcement to ensure cyclistscomply with traffic signals is needed.

Traffic signals at large junctions cateringfor NMVs should specifically consider theclearance times taken for the vehicles to travelbetween the arms of the junction. Appropriatesignal phasing and timings will help tominimize conflicts.

(iii) Roundabouts

Cyclists are particularly vulnerable atroundabouts and accidents can easily occur due

plate 4.14

Fig. 4.3Fig. 4.4

plate 4.15

Plate 4.14:Separate trafficsignals for cyclists ata junction in Beijing,PRC.

Plate 4.15 (far right):An advance stop linefor cyclists at asignal-controlledjunction in MarketHarborough, UK.

Figure 4.3:A right turn before thejunction (G turn)

Figure 4.4 (far right):A right turn after thejunction (Q turn)

37


to the high levels of conflicting movements andspeeds. Careful geometric design ofroundabouts is required to help reduce approachspeeds. Research in developed countries has ledto some special facilities to aid cyclists atroundabouts.

These measures are split into three types andinvolve either mixed flows on the roundaboutitself, some physical segregation on parts ofthe roundabout, or a separate track for cyclistsaround the outside of the roundabout. Mixedflow roundabouts involve small central islandsand single-lane approach roads with single-lanecirculatory movement, while segregationalmeasures reduce the risk of NMVs being hitwhile on the roundabout by vehicles enteringthe system.

Recently, moveable barriers have been pro-vided at the Sonargaon Roundabout in Dhaka,Bangladesh, although their effectiveness is notyet known.

(iv) Grade-separated crossings

Some footbridges in Indonesia and in anumber of cities in the PRC have been providedwith a central wheeling ramp. These rampswere initially intended to aid cyclists push theirmachine along, although in Indonesia, motor-cyclists were observed actually riding up anddown the ramps (see Plate 5.9).

Footbridges have a lower construction costthan underpasses but are less ideal for jointcycle and pedestrian usage, generally beingnarrower than underpasses, thus increasing thepotential for conflict. Also, the rider must dis-mount from the bicycle and wheel the machineup and down a smooth wheeling ramp — anunattractive procedure.

Underpasses provide grade-separated cross-ings of carriageways that would otherwise betoo dangerous or inappropriate to cross at grade.To be well-used, underpasses, as with foot-bridges, need to be conveniently sited to at-

tract high pedestrian and cyclist use.Dimensions of underpasses are important inallowing cyclists access to them and wherepossible, pedestrians and NMVs should besegregated in order to achieve safe access forboth. Approach ramps should not be so steepthat they encourage high speeds through theunderpasses or to make the exit too strenuous.Underpasses with specific NMV provision arerare in the region.

c) Parking facilities

An essential part of the cycle network is theprovision of cycle parking facilities to providesafe storage of the machine and reduce thethreat of theft or vandalism.

Suitable facilities that fulfil these require-ments involve the provision of an adequatenumber of cycle stands, allow either self-sup-plied or cycle stand provided lockingmechanisms, and are located at major trafficgenerators, as in Surabaya, Indonesia(PADECO Co., Ltd.). Attendants also reducethe threat of theft. In the Netherlands and otherdeveloped countries, such stands are increas-ingly being provided with shelter from theweather. Facilities for cycle parking exist inmost Asian cities where high volumes ofcyclists occur (for example, cities in the PRCand Viet Nam).

Official waiting areas for cycle-rickshawsare rare and many clusters of cycle-rickshawscan be seen to be operating from various streetcorners and major junctions in Asian countries.

Plate 4.16 shows rickshaws waiting at theedge of the carriageway, reducing the effectivecarriageway width and hence the road capacity.A further problem associated with the lack ofsuitable parking arrangements is shown in Plate4.17, where pedal cycles parked on the foot-way block pedestrian movement and force pe-destrians to walk in the road.


Plate 4.16:Rickshaws parked onthe side of the streetdue to a lack ofsuitable waiting baysin Yogyakarta,Indonesia.

Plate 4.17 (far right):Pedal cycles parked onthe footway blockingpedestrian access inHanoi, Viet Nam.

38


4.4 Integrated Programs

a) Enforcement

The enforcement of traffic regulations gov-erning all road users and vehicles is essentialfor the safety of pedal cyclists and other NMVs.Regulations relating to motor vehicles that canincrease road safety include the enforcementof speed restrictions and waiting restrictions inthe vicinity of on-road cycle lanes. Similarly,NMV users have a responsibility to acknowl-edge and obey the various traffic laws. For ex-ample, NMVs should obey traffic signals, havelane discipline, and not encroach on to the foot-way. They should be prosecuted if they do notobey these laws.

Furthermore, it is important that pedal cyclesor any NMVs are as fit for the road as any motorvehicle. Hence equipment such as reflectors,headlights, and braking systems should befitted and in good working order. There is noreason that these should not be a legalrequirement and checked for compliance.However, this is rarely done in practice.

Traffic police must treat NMVs in the sameway as other road users. They should have thesame rights to road space, but equally, theyshould have the same obligations to obey traf-fic laws.

Traffic regulations are not self-enforcingand it is suggested that where possible,measures be provided to ease enforcement, forexample, by preventing motor vehicle accessto cycle paths through bollards across theentrance and exit points.

b) Education, training and publicity

Although road safety education is part ofthe national school curriculum in many countriesin the region, little attention or training is givento NMV users, who will often ride their ma-chines in mixed traffic streams at a very youngage with little or no knowledge of trafficregulations.

Training to increase knowledge of trafficregulations and improve cycling skills shouldbe encouraged and should be conducted byskilled instructors. Training should includeinformation on the correct positioning of vehi-cles at junctions relative to the required exitarm and on the dangers of traveling along roadlinks in the wrong direction. Publicity cam-paigns should be provided to educate cyclistsand motor vehicle drivers about NMV safetyand on-road dangers. The highway code, if itexists, should also contain a section on pedalcycle and NMV considerations.

39


Motorcycles are an important part of the traffic fleet in many countries of the Asianand Pacific region and in several countries they are the dominant vehicles. Despitethis importance, accident data on involvement of motorcycles in road accidents in the

Asian and Pacific region is sparse. Only a few countries were able to provide the necessarybreakdowns into vehicle types involved in accidents and casualties.

In addition, despite their relative importance in the traffic stream, few efforts have been madeto provide special facilities to cater for the needs of motorcycles. In comparison withpedestrians, little research has been conducted into the needs of motorcycles and their roadsafety problems.

The authors’ efforts in examining the needs, problems, and facilities of motorcycles were,therefore, characterized by lack of data and research, and the absence (apart from Malaysia) ofefforts to provide safety facilities for these vehicles.

5 MOTORCYCLES

5.1 Introduction

5.2 Motorcycle Accident and Vehicle Data

Motorcycles are involved in a substantialshare of fatal and injury accidents in severalcountries in the region (see Tables 1.1 and 1.2).

Table 5.1 shows the extent of this involvementin fatal and injury accidents. Comparatively,motorcyclists are involved in a greater share ofinjury accidents than fatal accidents and in threeof the countries, motorcyclists are involved inmore than half of the fatal accidents.

Table 5.1: Percentage of Fatal and Injury Accidents Involving Motorcycles

Subregion Country Fatal accidents (percent) Injury accidents (percent)

NIEs Hong Kong, China 10 17 Republic of Korea 7 6Singapore 51 69Taipei,China 54 62

Central Asia People’s Republic of 10(1)

na China (1994)


74(1)


PDMCs Fiji(1)

na

Developed New Zealand 16 14

na Data not available.(1) Data relates to casualties.Source: RETA project data.

40


Table 5.2 shows the percentage of registeredvehicles that are motorcycles in various coun-tries in the region, along with the motorcyclefleet growth or decline between 1989 and 1993.An important point is that of the countries, suchas Indonesia and the Philippines , that haveprovided motorcycle fleet information, thereis no data on their involvement in accidents.

Countries with high motorcycle accidentinvolvement rates, such as Malaysia andTaipei,China, also have large motorcycle fleetsrelative to the total number of motor vehicles.It should be noted that except for Hong Kong,China, the motorcycle share of motor vehicletraffic in the NIEs is decreasing. However, withthe exception of Singapore, the motorcyclefleet has actually increased in size, indicatingthat the total number of motor vehicles has alsorisen but motorcycles are increasing lessrapidly. In the developed countries, both thenumber and percentage of motorcycles, rela-tive to the total motor vehicle fleet, havedecreased steadily in recent years.

Motorcycles are a much more affordablevehicle type compared to other motor vehiclesand this may lead to motorization levels increas-

ing at an even higher rate than found in devel-oped countries. In the most motorized countryin the Asian and Pacific region — Taipei,China— more than 70 percent of its motor vehiclesare motorcycles. Apart from the Philippines ,two- and three-wheeled motor vehicles accountfor about half of all motor vehicles in the SouthAsia and Southeast Asia subregions. In India,Indonesia, and Thailand, they represent twothirds of all motor vehicles. Motor vehiclegrowth is expected to increase even faster thanat present and with motorcycles playing a domi-nant role, this low-cost and vulnerable motor-ized vehicle could conceivably accelerate boththe motorization and death rates in the Asianand Pacific region.

a) NIEs

Hong Kong, China10 percent fatal accidents,17 percent injury accidents

Thirty-five fatal and 2,582 injury accidentsinvolving motorcycles occurred in Hong Kong,

Table 5.2: Percentage of Registered Motor Vehicles that are Motorcycles and their Rate of Growth

Subregion Country Percentage of motor Percentage change invehicles that are registered motorcycles

motorcycles (1993) (1989-1993)

NIEs Hong Kong, China 5.3 +41Republic of Korea 23.6 +63Singapore 20.5 -1Taipei,China 73.7 +44

Central Asia People’s Republic 36.8 +139 of China

Southeast Asia Indonesia 68.4 +22Lao PDR na +80Malaysia 55.8 +30Philippines 25.8 +65Thailand 65.6 +75Viet Nam na +175

South Asia Bangladesh na +48India 67.3 +60Pakistan 48.1 +32Sri Lanka na +68

(1)

Pacific Tonga na +45

Developed Australia 2.8 -8countries Japan 21.0 -10

New Zealand na -38

na Data not available.(1) Growth between 1989 and 1992.Source: RETA project data.

41


China in 1993, making up 10 percent and 17percent, respectively, of the total accidents. Noobvious trend in accidents from 1989 can beseen, as fatal accidents involving motorcyclesincreased at the same rate as all fatal accidents.Although the number of motorcycles in HongKong, China has increased over time, they havenot increased in number as quickly as motorvehicles and thus form a lower percentage shareof vehicles.

Republic of Korea7 percent fatal accidents,6 percent injury accidents

In 1993, there were 673 fatal accidents and13,912 injury accidents involving motorcyclesin the Republic of Korea. There has been a de-creasing trend in the number and share of bothtypes of accident since 1989. The Republic ofKorea, similar to Hong Kong, China, has agrowing motorcycle fleet but a decreasing pro-portion of the total number of motor vehicles.

Singapore51 percent fatal accidents,69 percent injury accidents

Singapore had 123 fatal and 3,840 injuryaccidents involving motorcycles in 1993, mak-ing up 51 percent and 69 percent, respectively,of the total fatal and injury accidents. Since1990, fatal accidents involving motorcycleshave been increasing and are now at their high-est level ever. This increase in fatal accidentsis even more worrying when it is noted thatsince 1990, the motorcycle fleet in Singaporehas been declining.

Taipei,China54 percent fatal accidents,62 percent injury accidents

Taipei,China recorded 1,125 fatal accidentsand 389 injury accidents involving motorcy-cles in 1993, working out as 54 percent and62 percent, respectively, of the total accidents.Taipei,China is seeing a downward trend infatal accidents, with those involving motorcy-cles decreasing more rapidly. It has an in-creasing motorcycle fleet size that is increas-ing less rapidly than other types of motor ve-hicle.

b) Central Asia

The People’s Republic of China10 percent fatalities

The only relevant motorcycle data availablefor this subregion are for the PRC. Motorcy-cles were involved in 13 percent of all motorvehicle deaths and 10 percent of all road deathsin 1994. The PRC has also seen the number ofregistered motorcycles treble in size between1989 and 1994, with an increase in the percent-age share of all motor vehicles increasing from27 percent to 40 percent. Table 5.3 shows thisincrease.

c) Southeast Asia

Of the seven Southeast Asian countries,some motorcycle-related data were availablefrom all the countries, except Cambodia.

Table 5.3: Registered Motorcycles in the PRC

Year Motorcycles Motor vehicles Motorcycles as percentage (10,000) (10,000) of total motor vehicles

1990 359.33 1318.53 27

1990 421.28 1,476.26 29

1991 505.15 1,657.66 30

1992 647.49 1,945.03 33

1993 858.79 2,331.64 37

1994 1,093.81 2,735.60 40

42



Indonesia69 percent of motor vehicles aremotorcycles

Registered motorcycles in Indonesia havesteadily increased in number from 5.7 millionin 1989 to 7.75 million in 1994 and make up69 percent of the total motor vehicle fleet. Thispercentage share of the motor vehicle fleet hasremained fairly constant since 1980.

Lao PDRMotorcycle fleet doubled in five years

Motorcycle numbers in Lao PDR have morethan doubled between 1989 and 1994. In 1994there were 109,490 registered motorcycles.

Malaysia57 percent fatal accidents,74 percent injury accidents

More than half of the total road accidentdeaths and almost three quarters of the totalinjuries in 1994 in Malaysia involved motor-cycles. The number of deaths and injuries in-volving motorcycles has increased since 1989,along with their proportional involvement inaccidents, from 1,608 (43 percent) deaths in1989 to 2,946 (57 percent) in 1994 and from13,386 (51 percent) injuries in 1989 to 31,957(74 percent) in 1994. During this period, themotorcycle fleet increased by 40 percent to 3.7million registered motorcycles, which make upmore than 55 percent of the total motor vehiclefleet in Malaysia.

This percentage share of all motor vehicleshas remained relatively constant throughoutthe 1989-1993 period.

Malaysia has about 4 million registeredmotorcycles and their proportion on the roadvaries between 13 percent and 75 percent, de-pending on location. The motorcycle propor-tion of registered vehicles ranges from 36.1 per-cent in Kuala Lumpur to 67.1 percent in Perak.

PhilippinesMotorcycle fleet doubled in six years and isnow 25 percent of motor vehicle fleet

The number of motorcycles and its propor-tion within the motor vehicle fleet in the Phil-ippines has increased to more than half a mil-lion in 1993, representing a doubling in fleetsize since 1987.

More than a quarter of registered motorvehicles in 1993 was motorcycles.

ThailandMotorcycle fleet doubled in five years

In 1993, more than 7.25 million motorcy-cles were registered in Thailand with thefleet size growing by almost 1 million from1992.

The actual number of registered motor-cycles doubled between 1988 and 1993. Mo-torcycles represented 64 percent of the totalmotor vehicle fleet in 1988 and 66 percentin 1993.

Viet NamMotorcycle fleet trebled in four years

The number of registered motorcycles inViet Nam has trebled from under 1 million in1989 to more than 3 million in 1993. A TRRarticle claims that motorcycles were involvedin more than 60 percent of road accidents inHanoi (see Bell and Kuranami, 1994).

d) South Asia

BangladeshMotorcycle fleet increased by40 percent in five years

The number of registered motorcycles inBangladesh has increased from 140,100 in 1989to more than 200,000 in 1994, an increase ofmore than 40 percent.

India67 percent of motor vehicles are motorcycles

There has been a constant increase in thesize of the registered motorcycle fleet in Indiasince 1980, with the fleet growing at a fasterrate than other motor vehicles.

In 1993, India had about 17 millionregistered motorcycles, comprising more than67 percent of total motor vehicles.

Pakistan48 percent of motor vehicles aremotorcycles

The number of motorcycles in Pakistan,along with their proportion of total motorvehicles, fluctuated between 1988 and 1993.

However, in 1993 more than 1.5 million mo-torcycles comprised 48 percent of the totalmotor vehicle fleet.

43


Sri Lanka16 percent fatal accidents,26 percent injury accidents

The number of fatal accidents involvingmotorcycles in 1993 in Sri Lanka was 208(16 percent of total fatal accidents) while therewere 3,610 injury accidents involving motor-cycles (26 percent of total injury accidents).Between 1989 and 1993, motorcycle involve-ment in fatal accidents fluctuated while injuryaccidents showed an overall increase.

e) PDMCs

Very small percent of motorcycleaccidents

Of the 11 nations in the Pacific islands, onlyFiji, Samoa, and Tonga could provide anyrecent data involving motorcycles. A smallnumber of fatal and injury accidents involvingmotorcycles has been recorded in the threecountries. Samoa had 11 injury accidentsinvolving motorcycles, which comprised 9 per-cent of the total injury accidents.

5.3 Motorcycle Facilities

Although few physical engineering facili-ties to improve motorcycle safety exist, somemeasures have been identified and are consid-ered important. Motorcycle safety can be in-creased with the separation of two- and three-wheeled motorcycles from large, high-speedvehicles.

Furthermore, motorcyclists will benefit fromspeed reduction measures where there ismixed traffic (see section on traffic calmingand speed reduction measures in Chapter 6).

a) Road links

Increased safety can be achieved by theseparation of motorcycles from other motor ve-hicles. This segregation can take one of twoforms. Either exclusive motorcycle lanes canbe created, as in Malaysia. These lanes areseparated from the main carriageway by aphysical median; or joint motorcycle and SMVlanes can be provided, as in Kuala Lumpur andViet Nam. These joint lanes provide routes thatpedal cyclists and other nonmotorized vehiclescan also use.

The exclusive motorcycle lane in Malaysiais 14 km long and has led to a recorded reductionin accidents of 27 percent with a benefit to costratio of constructing the lane valued at aboutthree. A subsequent extension constructed in1992 is estimated to have reduced motorcycleaccidents by 34 percent along the section of roadconcerned. Consequently, the exclusive motor-cycle lanes in Malaysia appear to have beensuccessful and may be worth considering insimilar circumstances where high motorcycleflows are found along a corridor.

During the ADB/UN/ESCAP seminar inBangkok (1996), some delegates expressed theview that motorcycles and pedal cycles/NMVsshould be kept separate due to differences inspeeds. This occurs on some roads in the PRCwhere such segregated lanes were observed tobe operating (see Plate 5.1). In this instance,guardrails separate the motorcycles from a jointpedal cycle and pedestrian path, while a raisedmedian segregates the whole path from the maincarriageway.

Plates 5.2 to 5.5 show examples of sharedmotorcycle and pedal cycle lanes in KualaLumpur, Malaysia, and Viet Nam. Differencesin lane width are clearly evident from the pho-tographs; sufficient lane widths can be deter-mined from traffic counts and forecasted flowsof vehicle modes. Clear road signs for the jointuse of the lane is also evident in Plate 5.5. Theuse of these slow moving lanes to accommo-date motorcycles and pedal cycles was alsobeing proposed as part of the Yogyakarta Mas-ter Plan in Indonesia. The plan included de-tails of separate lanes of about 3.5 m to 4.5 mwide on primary collector roads.

plate 5.1

Plate 5.1:Segregatedmotorcycle lane andshared pedestrian andpedal cycle lane in thePRC.

44


plate 5.3plate 5.2

Despite the provision of separate SMVlanes, in certain circumstances, the shared useby NMVs and motorcycles is generally notallowed and motorcycles must usually use themain carriageway, as shown in Plates 5.6 and5.7.

Some motorcycle and motor vehicle sepa-ration can be achieved by allowing the shareduse of bus lanes. However, full considerationof the traffic flows of both types of vehicle isimportant — shared use at specified times of

plate 5.5plate 5.4

the day could be a possible acceptable measure.Care must be taken not to encourage thesharing of all facilities such as pedal cycle

plate 5.7plate 5.6

plate 5.9plate 5.8

Plate 5.2:Motorcycle and pedalcycle SMV Lane in HoChi Minh City, VietNam.

Plate 5.3 (far right):SMV Lane at ajunction in Ho ChiMinh City.

Plate 5.4:Motorcycle and pedalcycle shared lanes inKuala Lumpur,Malaysia.

Plate 5.5 (far right):Motorcycle and pedalcycle shared lanes inKuala Lumpur. Noticethe explicit signs.

Plate 5.6:Motorcyclist usingmain carriageway inthe PRC. Note thesegregated cycle lane.

Plate 5.7 (far right):Motorcyclists forcedto use the maincarriageway in VietNam.

Plate 5.8:Posts placed across aroad prevent four-wheeled vehicleaccess in Bandung,Indonesia.

Plate 5.9 (far right):Motorcyclist using acentral wheeling rampintended for pedalcyclists on afootbridge in Jakarta,Indonesia.

45


measures at junctionsor even on foot-bridges, due to thedifferences in res-pective vehicle speeds.Plate 5.9 illustratesthe misuse of a centralcycle wheeling rampby a motorcycle inJakarta, Indonesia.

Alternative roadlink measures to pre-

vent four-wheeled vehicle access were also ob-served in Bandung, Indonesia , where posts atthe entry/exit points were used (see Plate 5.8).

b) Road surface and trafficengineering

Several Australian safety specialists havehighlighted the need to give greaterconsideration to street maintenance and traf-fic engineering as these can have a seriouseffect on motorcycle safety.

Australia’s AustRoads was planning to pub-lish Traffic Engineering and Motorcycle RoadSafety and Australian safety specialists suggestthe following areas for consideration regardingtraffic engineering and motorcycle safety.

i) Highway geometry: Adverse camberis especially hazardous for motorcyclesbecause of the need to lean even fur-ther for balance. Where this occurs withuneven road surfaces or localized lowfriction areas of pavement it can be dan-gerous for motorcyclist safety.

ii) Road surface:

1 Manhole covers and steel plates usedto cover roadworks are often ex-tremely slippery. The application ofan abrasive coating reduces the dan-ger.

2 Speed bumps and traffic calmingmeasures of poor design can be dan-gerous to motorcycles. Designs suchas those advocated by TRL and theAustralian Road Research Boardshould be used.

3 Loose gravel gets onto corners ofsealed roads when shoulders are un-sealed and can be dangerous to mo-torcyclists. At intersections on roads

that have unsealed shoulders, suchproblems can be overcome by seal-ing to a curved section of kerb andgutter.

iii) Intersections: Motorcycles are fre-quently masked (hidden) by other ve-hicles or not noticed, so fully control-led turns should be used at signalizedintersections, wherever possible, toavoid motorcycle accidents. Sharplyflared intersections and smallroundabouts with painted central is-lands can be dangerous for motorcycles.Pavement loop detectors are sometimestoo insensitive (due to the light weightof motorcycles) and can fail to detectmotorcycles or allow them to clear theintersection before the lights change.

iv) Road markings: Variations in skid re-sistance from road markings (for exam-ple, painted arrows, lane and edge lines,or zebra crossings) when wet or oncurves can be dangerous for motorcy-clists. Raised markers used to delineatelane and edge lines can also cause prob-lems, especially at curves.

v) Safety barriers: Collision with a nor-mal w-beam (Armco) steel barrier canbe dangerous for motorcyclists as therider can be catapulted over or underthe barrier and reflectors on the barri-ers can cause injury to a rider slidingalong the barrier. Wire rope barriers areparticularly dangerous for motorcyclistsand support posts and end treatmentsof all types of barriers can cause severeinjury to motorcyclists. From a motor-cyclist’s viewpoint, the concrete taperedshape (New Jersey) barrier is probablythe least damaging in the event of anaccident.

vi) Roadside furniture and objects:Roadside vertical kerbing can causeadditional dangers to motorcyclists slid-ing along a road after an accident un-less it is of rounded semimountable de-sign. Roadside poles, if placed too closeto the kerb at corners, can be danger-ous as motorcyclists have to lean overas they go round a bend.

These and other potential dangers to motor-cyclists are discussed in more detail in theAustRoads document, Traffic Engineering and

plate 5.10

Plate 5.10:Motorcycle parkingsigns in Yogyakarta,Indonesia.

46


Motor Cycle Road Safety. What is clear is thatparticular care needs to be given to the designof road and traffic engineering facilities where alarge number of motorcyclists can be expectedin the traffic stream. Although such measureswill not completely eliminate motorcycleaccidents, they will minimize their occurrenceand reduce their severity when they do occur.

c) Parking facilities

Parking facilities form an essential part ofoverall motorcycle requirements. Facilitiesshould be well signposted and be able to copewith the expected motorcycle numbers. Wherenumbers of vehicles are expected to be highand where motorcycles form the largest shareof all modes in traffic, special parking lots ex-clusively for motorcycles could be considered.Signs for motorcycle parking facilities are es-sential, as can be seen in Plate 5.10. Plates 5.11and 5.12 clearly illustrate that a lack of adequateparking can cause blockages of footways androads. Parking lots with an attendant presentwill help reduce the threat of theft.

5.4 Integrated Programs

Many motorcycle road safety initiatives relynot on physical engineering measures but onassociated measures such as road safety edu-cation and suitable traffic regulations. Thesemeasures in the past have typically been basedupon rider training and helmet legislation.

a) Enforcement

Motorcycle helmet wearing can, if correctlylegislated and enforced, reduce the severity ofmotorcycle accidents. Legislation must provide

for the compulsory wearing of helmets whileriding a motorcycle of any size and speed; thismust be backed by enforcement. However,helmet design and construction must relate toa suitable minimum legal standard that willoffer the wearer an appropriate amount ofprotection in the event of an accident. Trafficpolice using motorcycles must be seen to wearthe correct helmet type, have the helmet fas-tened, and to enforce their correct usage. Hel-mets should also be worn by pillion passen-gers.

Regulations concerning the size and speedof motorcycles should also be considered, alongwith the use of graded licenses based upon ageand experience.

b) Education, training, and publicity

The training of motorcycle riders, as withany motor vehicle driver, is an important safetyrequirement and should be carried out bytrained and qualified instructors. Trainingshould cover information not only based on thedevelopment of driving skills and driver respon-sibilities, but also on safety factors such as therisk of not wearing an approved safety helmet.An official on-road riding test should be car-ried out in order to check the motorcyclists’competence and riding skills. Plates 5.13 and5.14 show motorcycle training and testing inSingapore. Information from the 1994Malaysian Road Accident Statistical Reportnotes that motorcycle riders were at fault in 50percent of the accidents they were involved inand that more than 50 percent of the motorcy-cle riders involved in accidents were aged 16to 25. This highlights the need for suitable train-ing of young riders.

Potential dangers associated with motorcy-cle riding should be publicized (see Plate 5.15)


Plate 5.11:Motorcycle parking onthe footway blockingpedestrian access.

Plate 5.12 (far right):Motorcycle parkingblocking a lane in thecarriageway.

47


along with enforcement and legislationmeasures.

Safety publicity should be sufficiently pow-erful to counteract other influences presentedin the mass media, including speed and glam-our images that are used to advertise powerfulmotorcycles.

One of the most effective areas for educa-tion and publicity is to ensure the conspicuityof motorcyclists. Many accidents occur becauseother vehicle drivers do not see the motorcy-clists. Consequently, efforts need to be madeto encourage motorcyclists to increase theirconspicuity.

This can be done in three main ways:

i) ride with headlights on (dipped) at alltimes (day and night) as this makes amotorcycle much more conspicuous toother road users;

ii) riders should wear light colored cloth-ing and preferably reflective jacketswhile riding motorcycles; and

iii) use reflective strips on front and rearmudguards and along sides of the mo-torcycle to make it more conspicuous atnight.


plate 5.15

Plates 5.13 and 5.14:Motorcycle drivertraining and testing inSingapore.

Plate 5.15:Signs on the side ofthe road inYogyakarta,Indonesia, promotingthe use of motorcyclehelmets.

48


Safe planning by separating vehicles and preventing the interaction of various roadusers can reduce the number and severity of conflicts that occur between motorvehicles and VRUs, high-speed traffic and low-speed traffic, heavy goods vehicles, and

other road users. The planning process influences safety through consideration of variouscomponents, such as land use planning, network design, and the road hierarchy.

Although the safe planning of roads is the primary step towards the prevention ofconflicts, the safe design of roads through standards and specifications is also important. Thedesign and suitable provision of the facilities relevant to VRU safety for each group arediscussed in the preceeding chapters. It is intended that this section will look at those factorsthat are relevant to all three VRU groups together.

6 RECENT DEVELOPMENTS

6.1 Introduction

6.2 Land Use Planning

The control of land use and the traffic asso-ciated with the different areas of use is essen-tial to successful planning. Incompatible traf-fic patterns and vehicle compositions fromvarious land use sources should be separatedin order to minimize potential conflicts. Hence,zonal planning should be considered to reducetrip lengths and to prevent VRU interactionswith high speed and heavy volumes of traffic.Access control is also important in preventinginappropriate traffic mixing with VRUs.

Of particular im-portance to VRUs isadoption of the princi-ple of locating trafficgenerating facilitiesadjacent to a roadsuited to the type oftraffic expected, suchas the provision ofcoherent and linkedcycle paths and foot-ways close to a school.Furthermore, conflictscan be avoided by

preventing the development of linked land useon either side of main roads; for example,schools and housing estates on opposite sidesof a primary distributor road.

6.3 Road Hierarchy

The roads within a network should be de-fined and classified according to their func-tion, with the network traditionally split intolevels according to the primary use of the roadand thus the associated traffic. VRUs shouldbe considered within the planning and designof the network and road hierarchy in order thattheir own network of facilities be provided.Hence, coherent, direct, and safe routes forVRUs can be provided through measures suchas linked footways, cycle lanes, and road cross-ing facilities.

The identification of VRU networks can al-low the planning and provision of appropriatefacilities where the motor vehicle and VRUnetworks interact. For instance, underpassesand footbridges are not necessary for local ac-cess roads; similarly no special provisions areneeded for pedal cycles and other nonmotorized

plate 6.1

Plate 6.1:Well-planned roadnetwork hierarchy canreduce conflicts andimprove safety.

49


vehicles on low-flow and low-speed roads. Ta-ble 6.1 shows pedestrian and pedal cycle move-ments in relation to the typical traffic activityon various categories of road.

Traffic management through the creationof a suitable road hierarchy and the networkdesign can have positive effects on safety bylimiting traffic levels and speeds on certaintypes of road, thus benefiting VRUs. Trafficmanagement should help pedestrians and roadvehicles remain on their respective facilitiesand to minimize their interaction. Factors suchas improved traffic efficiency through reducedcongestion levels and/or an improvement in theenvironment are also byproducts of goodtraffic management through suitable road hi-erarchy planning and design.

Appropriate planning of the road hierarchycan lead to the maximizing of priority to groupssuch as pedestrians and cyclists over motor ve-hicles. This objective is an environmental con-cern that restricts motor vehicle access to pro-mote easier pedestrian and cycle use as well asgeneral safety (see Table 6.1).

6.4 Design Standards

Design standards for VRUs should reflectthe needs of each group in providing safe fa-cilities for the movement of people and goodsin order to reduce potential conflicts. However,for a facility to be fully utilized, such measuresshould be attractive to potential users in termsof the directness of the route. Facilities need to

be coordinated and cohesive in order to maxi-mize their potential benefit to all users whilebalancing the safety requirements.

The volume of VRUs within the overall traf-fic is an important consideration for designguidelines. In the case of NMV facilities, aseparate lane for sole use of NMVs requires awidth related not only to the volume of NMVtraffic but also the expected NMV trafficcomposition.

However, it is suggested that the speed, flow,and composition of motor vehicle traffic on theroad are the main factors in defining the needand type of cycle lane. Pedestrian volumes,however, are important to the level of facilitiesprovided. Footways should cater for theexpected pedestrian numbers along a route andbe of sufficient width in order to preventpedestrians having to walk on the carriageway,increasing the potential for conflict. A prob-lem with wide footway provision is theencroachment of squatter areas, which can cre-ate hazardous conditions for all road users. Inorder to provide suitable facilities for VRUs, itis necessary to determine potential usage. Thiscan be achieved only through carrying out pe-destrian and traffic counts.

This volume is not intended as a designmanual and although Chapters 3 to 5 providegeneral information on standards and capacityof different facilities, no attempt has been madeto recommend specific guidelines relating totraffic volumes or other such design require-ments. This document seeks only to identifyfacilities in use in the region, to indicate the

Table 6.1: Pedestrian and Pedal Cycle Activity with Respect to Road Category

Road category Main activity Pedestrian activity Pedal cycle activity

Pedestrian street Shopping, walking Predominant Considerable, speed limitedby pedestrians

Access road Walking, vehicle Considerable Considerable, crossing,access, deliveries freedom weaving, stopping

Local distributor Start/finish of vehicle Controlled Considerable, bicycle lane,trips, bus stops not physically segregated,

no crossing facilities

District Through traffic, Minimum positive Considerable, segregateddistributor public transport safety measures cycle lane, crossing

facilities

Primary Fast moving, through No safety measures — Controlled, segregateddistributor traffic segregation cycle lane.

Source: MVA, Beijing Traffic Study, 1986.

50


forms that they can take and, where possible,to comment on their perceived effectiveness.

6.5 Traffic Counts

Traffic counts along with accident record-ing (see below) provide the basis for designingtraffic systems that cater for all road users. Forexample, facilities for pedal cyclists and otherNMVs may depend not only on the number ofpedal cycles/NMVs, but also upon the numberof motor vehicles and how the two categoriesinteract. This relates to the defining of the roadhierarchy and to traffic flows within various landuse zones. Traffic volumes will be highest at ornear major generators and this should beconsidered during the planning of traffic counts.

Data in the form of vehicle and pedestriancounts can provide details of existing trafficmovement, how this flow has changed over aperiod of time, and how it is likely to change inthe future. This all helps in the design of ad-equate facilities to assist road users. As hasbeen shown, only limited data exists for NMVtraffic while pedestrian traffic flow is almostnever counted. An informal survey carried outat the ADB/UN/ESCAP conference in Bangkokrevealed that a few pedestrian counts have beencarried out in Yogyakarta, Indonesia; Karachi,Pakistan; Manila, Philippines; and Singapore.Data is available for motorcycles because oftheir place in mainstream traffic, as well as therequirement of registering vehicles, which pro-vides some details of ownership levels.

6.6 Accident Data Systems

Accident data provide the objective andscientific basis for determining not only theextent and type of safety problem experienced

but also the kinds of remedial measures mostlikely to be effective. Without the collection ofaccident figures, the safety of a nation’s roadnetwork cannot be assessed and monitored. Adatabase consisting of relevant accident infor-mation allows an understanding of the circum-stances and patterns of accidents by means ofanalysis. To complete the process, informationfrom the analysis system must be disseminatedto relevant organizations.

Traffic police are most ideally placed torecord and manage accident data, which shouldinclude general accident details such as roadtype, environmental features, and location, aswell as the vehicle/driver details and casualtydetails. Within this data collection, specific fac-tors should be included, such as pedestrian ac-tion and location. A common accident micro-computer-based database in use is TRL’s MAAP.At present in the Asian and Pacific region, it isbeing used in Fiji, Malaysia, Papua New Guinea,and Philippines , and in certain areas ofBangladesh, India, Indonesia, and Nepal.

A good accident reporting system is also ableto investigate hazardous locations and identifyany patterns or similarities between accidents.For example, MAAP has recently been intro-duced in Dhaka, Bangladesh, and despite ananalysis period limited to only six months, onelocation in Dhaka stood out for its accident risk.On one of the main approach roads to the city, 17road accident deaths (of which 15 were pedestri-ans) had occurred at the junction.

After identifying what could, withjustification, be called the world’s worst blackspot, the accident analysis program can alsocheck for similarities in accident timings, vehiclemaneuvers, and other factors such as vehicletypes involved.

The accident statistics presented in the pre-vious chapters give some indication of the lowpriority being given to VRU safety in the Asianand Pacific region. Many countries appear notto be monitoring VRU accidents, or at least notpublicizing data findings. According to the UN/ESCAP survey, while accident data are centrallystored in 13 of the 15 respondent countries,only seven countries use microcomputers foraccident data storage and four countries usemainframes.

Meanwhile, paper files (i.e., manual analysis)are still largely used in five countries, includingBangladesh, India, Indonesia, Nepal, and SriLanka (some of these countries, however, dohave the MAAP system in use on a pilot basisor running in parts of the country).

Figure 6.1:Primary collector roadthrough an urbanarea.

Source: Yogyakarta Urban Development Project, 1995.

51


6.7 Foreign Aid Projects

Until recently, foreign funded road and trans-port improvement projects traditionally had anarrow focus, being concerned primarily withmaximizing traffic efficiency along all roadlinks. However, in recent years there has beengrowing attention paid to the safety problem,environmental issues, and social impact createdduring the design of new transport infrastruc-ture. While VRU requirements have begun tobe addressed, the true impact of recent trans-port developments on VRUs safety and activ-ity is still largely unknown due to the lack ofinformation on accident involvement rates andthe volumes within the traffic flow. This lackof knowledge has contributed in the past to thelack of VRU consideration in the design andconstruction of many projects.

6.8 Traffic Calming and SpeedReduction Measures

In the past decade, two key road safety en-gineering ideas have been developed and im-plemented in Europe and other motorizedcountries. These two developments were traf-fic calming and safety audits, both of whichhave been credited with achieving accidentreductions, and improving environmental qual-ity and avoiding expensive construction modi-fication costs.

Traffic calming (where engineering meas-ures are used to reduce speed and volume ofmotor vehicle traffic) has proven successful indeveloped countries. Not only is the risk ofaccidents reduced but also the associated acci-dent severity. These benefits apply to all roaduser groups with the creation of a better envi-ronment, but are particularly beneficial to thesafety of pedestrians, pedal cycles, and othernonmotorized vehicles. Traffic calming meth-

ods involve the horizontal and vertical realign-ment of roads at selected locations using vari-ous methods, such as road humps and speedtables, chicanes, pinch points, and footwaywidening. Alterations can also be applied tojunctions by raising the whole junction layout,providing entry treatments across junctions andthe use of miniroundabouts to help controlspeeds. Such measures should be implementedas part of an area-wide set of measures.

The benefits from speed reduction have re-ceived much attention in recent years and aninternational review of speed and accidents con-cluded that, on average, each one mile per hour(mph) (1.6 km/h) reduction in speed wouldresult in a 7 percent reduction in fatal accidents.Another finding, which has been usedextensively in publicity campaigns, is the effectof speed on pedestrian casualty severity. At 20mph (32 km/h), 5 percent of pedestrians hit willdie compared to 85 percent at 40 mph (64 km/h)(while 95 percent of children hit at 40 mph willdie).

As traffic calming is meant to benefit a widerranging group by reducing vehicle speeds andvolumes, specific measures for NMVs are notnecessary, providing that their needs are fullyconsidered during the planning stage. For in-stance, where carriageway widths have beenreduced, the interaction of cyclists and motorvehicles in the remaining space is of prime con-cern. Facilities for cyclists to avoid or minimizethe effects of physical obstacles meant for high-speed vehicles should be considered and cateredfor wherever possible. The design andconstruction of traffic calming features can havea significant impact upon cycle use; this includesconstruction features such as smooth transitionbetween running surfaces. Inappropriatemeasures, such as rumble strips, should beavoided because of the discomfort caused toNMV users, or else the design should be suchthat it has minimal effect on NMV traffic.

Figure 6.3Figure 6.2:Collision diagram.

Figure 6.3 (far right):Accident factor grid.

52


plate 6.3

plate 6.4

plate 6.2

plate 6.5

Traffic calming is beginning to be appliedin the developing countries of the Asian andPacific region. In Nepal, village gateways andtwo armed roundabouts have been suggestedfor restraining speeds on highways traversingvillages. After being upgraded, highways gen-erally experience higher speeds and increasedaccidents. In order to control speeds where theroad passes through an inhabited area, villagegateways have been recommended with asignposted village archway and heavy vegeta-tion planted on the sides of the signs, and withthe added effect of rumble bars increasing inheight. If these measures do not prove effec-tive, then the more costly option of a two-armed

roundabout will be considered. Fiji has for sev-eral years been implementing traffic calmingmeasures on main roads as they pass throughvillages. Village gateways have been installedat several locations. Such measures have alsobeen proposed for Samoa.

In an attempt to bring consistency to thevarious local schemes, traffic calming designcriteria have been recently developed in SouthAfrica. The 16 criteria include traffic volumes,accident rate, public service vehicles, pedes-trian risk, 85th percentile speed, through traf-fic volumes, pedestrian volumes, parking/load-ing movements, schools/playgrounds, footway/verges, frontage/accesses spacing, sensitivearea, one- or two-way, stopping sight distance,gradient, and road type. The two pedestrianspecific factors are pedestrian risk, which is asubjective assessment of risk exposure, andpedestrian volumes, which is the total pedes-trian movement over a four-hour period alonga 150 m section. There are no specific NMVor motorcycle criteria as these are not majormodes in South Africa. The criteria areweighted with only the speed factor receivingthe highest weighting level (five) while pedes-trian risk is weighted at two and pedestrianvolumes at three. Guidance on safety-consciousplanning and design of road networks, andchecklists for safety checking of roads are pro-vided in the TRL document Towards SaferRoads in Developing Countries (see Ross, etal., 1991).

6.9 Safety Audit

Safety audit refers to the formal process ofreviewing road projects specifically to identifypotential or existing safety hazards in order toallow for timely proactive correction, especiallyfor pedestrians and other VRUs whose safetyneeds are often neglected in traditional trans-port planning and design.

Plate 6.2 (above):Traffic calming inTokyo, Japan.

Plate 6.3 (right):Town center trafficcalming in the UK.

Plate 6.4:Raised pedestriancrossing in Fiji.

Plate 6.5 :Gateway in Fiji.

53


Safety checking or safety auditing (whichis a more formalized process) offer opportu-nities to create safer roads for vulnerable roadusers.

Audits on road schemes can be carried outwith specific reference to any of the vulnerableroad user needs before a formal safety audit.In developed countries, audits of cycle routesare becoming the norm.

The elements that can be checked at variousstages for each VRU are shown in Table 6.2. Itshould be noted that these considerations arefactors within the overall full formal safetyaudit.

Nine of the 15 countries responding to theRoad Engineering/Safety Infrastructure sectionof the UN/ESCAP survey claimed to have con-ducted safety audits or safety checks beforeconstruction and the Republic of Korea re-ported that its safety experts commented onschemes, although not as part of formal roadsafety audits.

The extent to which these safety audits existin practice is not known. The only verifiedsafety audit procedures are from Fiji, Nepal(where a draft Road Safety Audit Manual hasrecently been produced and includes a sampleaudit), and Papua New Guinea, where safetyaudit procedures have been introduced as partof recently completed ADB-funded safetyprojects.

The Department of Roads in Nepal recentlybegan requiring all major new road projects toundergo a road safety audit. The key princi-ples of the road safety audit process in Nepalinclude:

• designing for all road users; and

• encouraging appropriate speeds andbehavior by design (Transport Engineer-ing and Safety Unit [TESU], 1996).

The first principle refers to the break withtraditional road engineering’s focus on pave-ment design and four-wheel motorized vehi-cles, and ensuring that all road users’ needs areconsidered in the design process. The secondVRU-related principle given is to ensure ap-propriate speed through good design. Forexample, when major national roads passthrough small communities or towns, speedsthat were perfectly acceptable on the rural sec-tions of the roads will no longer be appropri-ate. Thus, through design, drivers have to bemade to realize that they are entering a differ-ent “zone” where high speed is no longeracceptable. Use of such features as rumblestrips and gateway features as motorists ap-proach communities give a clear indication of achange in road character and this should re-sult in reduced speed as motorists pass throughthe community.

54


Table 6.2: Safety Audit Considerations for VRUs

Audit Vulnerable Road User Design Considerationsstage

Pedestrian Nonmotorized vehicle Motorcycle

Feasibility Impact of road improvements Impact of road improvements Scheme consistent within area of high activity. in area of high activity. adjacent road standards.Severance of existing route. Severance of existing route. Severance of existing two-Expected flows — along and Expected flow on/across wheel vehicle route.across the road. the road.Pedestrian provisions — Cycle lanes/crossings.footway and crossings. Accident data for similarAccident data for similar schemes.schemes. Junction numbers minimized

and of correct type.

Preliminary Establish sightline design Establish sightline design Suitability of separate lanesdesign standards. standards. — merging/diverging

Adequate footpath provision. Merging/diverging areas locations, sightlines,Crossing provisions at with motor vehicles. and speeds.major/minor junctions and Crossing facilities.traffic signals. Suitable junction types andSuitable crossing locations size considered re.and types. movements.

Adequate cycle pathprovision.

Final Signs/markings/street furniture Signs/markings/street Signs/markings/streetdesign — correct position, furniture — correct position, furniture — correct position,

conspicuity and visibility. conspicuity and visibility. conspicuity, and visibility.Adequate footway provision Landscaping effect on Landscaping effect onwith cohesive/direct routes. visibility. visibility.Pedestrian guardrail provision Correct materials specified. Advancing warning ofand location. Lighting requirements. crossings.Landscaping effect on Approach speeds of vehicles Turning conflicts atvisibility. at crossing locations. junctions minimized.Correct materials specified. Dropped kerbing. Motorcycle paths ofLighting requirements. Turning conflicts at junctions suitable design standard.Approach speeds of vehicles at minimized. Traffic regulations to allowcrossing locations. Roundabout approaches/ specified access.Dropped kerbing suitability. circulatory lanes.Zebra, pelican crossings and Overtaking at refugespedestrian phase at signal prevented.junctions — signal heads Cycle crossings provided.visible, central refuge, signal Adequate and suitable cycletimings, and banned turns, path provisions andpedestrian visibility to drivers. dimensions.Refuges — size, visibility, Traffic regulations to banlocation; overtaking prevented. waiting vehicles on lanes.

Pre- Adequate footway provisions. Cycle paths of adequate Motorcycle paths ofopening Footways connected — dimensions and linked. adequate dimensions and

direct, cohesive safe routes. Smooth transitions between fulfilling design standards.Signs and guardrails checked. riding surfaces. Smooth transitions betweenRoute should be walked, Route should be walked, riding surfaces.cycled, and driven along cycled, and driven along Route should be walked,during the day and night. during the day and night. cycled, and driven along

Publicize that path exists. during the day and night.

55


7 CONCLUSIONS

Although pedestrians are the largest casualty group involved in road accidents inmost Asian and Pacific countries, motorcycle accidents have become, or arerapidly becoming, a serious problem. Pedal cycles and other NMVs, however,

according to the national statistics of most countries, seem to contribute far less to roadaccidents than their presence in the traffic stream would suggest. Conflicting evidence,though, from Beijing, PRC, shows that they are involved in a high proportion of roadaccidents in that city. These three groups together are often involved in more than 75 percentof all fatal and injury accidents in much of the region.

Despite the extensive presence of pedestri-ans, pedal cycles, NMVs, and motorcycles inthe traffic stream, there is a considerable lackof information regarding the proportion andusage of these modes and their importance withrespect to road accidents in many countries. Thelack of comprehensive accident data systems,coupled with the underreporting of many acci-dents, prevents detailed analysis from being car-ried out and thus the most suitable remedialmeasures cannot necessarily be implemented,be they engineering, enforcement, or education.An efficient accident data system that collectsaccident details, such as VRU actions, age andsex, as well as the accident locations, needs tobe implemented as a matter of priority in coun-tries where it is missing.

Motor vehicle growth in the region is be-tween about 16 percent and 18 percent, and itis feasible that this rate could increase sincethe number of motorcycles is increasing morerapidly than other types of motor vehicle inmany countries. This situation is compoundedby many governments appearing to be activelydiscriminating against NMVs with policiesdesigned to discourage their use. The increasein the proportion of motorcycle traffic mayhave a significant detrimental impact on fu-ture accident rates.

In the Asian and Pacific region, there is asevere lack of priority and attention given toVRU movements, despite VRUs dominatingtravel patterns and casualty types. While some

consideration has been given to pedestrians’needs, especially as they cross roads and comeinto direct conflict with vehicle flows, NMVsand motorcycles are left to manage in the gen-eral traffic stream alongside larger and fastermotor vehicles. It may be unreasonable to ex-pect NMV policy or geometric design stand-ards to be radically revised so as to reflect amore balanced approach towards the differenttravel modes, but much more could be done toimprove road safety and facilities for suchgroups.

More attention should be given to VRUsthrough improved land use planning and roadhierarchy. This is justified both by their volumeas well as their vulnerability. A better approachto VRUs (especially pedestrians and NMVs)includes considering their needs at the earliestplanning stages instead of the traditional trafficengineering focus on four-wheel motor vehicletraffic. Land use planning can minimize triplengths and frequencies while a well-designedroad hierarchy would match trip types to roadclasses, thereby reducing the conflict betweensuch different modes as pedestrians, cyclists,private cars, and large commercial vehicles.

A practical and realistic approach to improv-ing VRU safety is needed. Road user condi-tions vary so much from that of developedcountries that many of the measures and as-sumptions used in the developed world do notapply for the developing Asian and Pacificregion. What has been learned from the review

56


of VRU safety presented here is the lack of anysimple and effective measures that are uni-formly successful across the region. Even fa-cilities tailored to local conditions can havedifficulties. For instance, raised zebra cross-ings may have insufficient impact on drivercompliance (even with high humps), 1.5 mguardrails may still be climbed, and the levelof enforcement needed to ensure proper use offacilities provided may well be missing.

None of the above changes or reduces theneed for improving VRU safety. It does, how-ever, support the case for the development oflow-cost measures that can be widely imple-mented and field-tested. It is recommended thatthe following strategy be considered for thedevelopment of VRU facilities and the im-provement of VRU safety.

Immediate Action

i. Scarce resources dictate that the prior-ity focus should be on known VRUproblem areas, such as junctions forNMVs and motorcyclists, and mid-linkcrossings for pedestrians. Conflict ar-eas need to be addressed in a realisticand pragmatic manner. For instance, nopriority should be assigned to VRU fa-cilities unless it can be enforced; i.e.,at-grade zebra crossings should not beused as they can offer only a mislead-ing sense of priority and a false senseof security unless there is a willingnessto enforce driver observance.

ii. On primary roads, NMV turns mayneed to be banned as part of a generaltraffic rule to ensure NMVs remain inthe slow lane and do not conflict withfaster motorized vehicle flows.However, in order to ensure a reason-able level of compliance, attractivetraffic management alternatives suchas G turns and Q turns need to beprovided to prevent NMVs frommaking the particular banned turn.

iii. VRU-friendly design should be soughtwith wider and flat-topped mediansused. Refuges can also be provided anddropped edge detail could be suitablefor NMV usage.

iv. As VRU safety is insufficiently cateredfor in traditional road engineering,safety audits should be used to double-

check all proposed schemes for theirimpacts on VRU safety and to ensurethat NMV traffic movements can beundertaken safely.

v. Traffic calming measures should be in-troduced in urban and rural locations toensure speed is at an appropriate levelfor the road, and its traffic and sur-rounding land use.

vi. Research should be undertaken to de-velop and assess facilities to improveVRU safety.

Longer-term Action

Over time, VRU consideration should beincorporated into the planning stage. A well-defined road hierarchy would assist, as VRUswould receive priority on the lower categoryroads and a NMV network could be developed.An approach based solely on improvement ofisolated black spots is insufficient and needsto be extended to development of safe networksfor NMV use.

While land use planning and a road clas-sification system will help minimize conflictsbetween VRUs and motorized four-wheeledvehicles, VRU movements and flows will con-tinue to dominate traffic flows and increasedfunding will be needed to provide VRU facili-ties.

Finally, two main points need to be stressed:

i. The trade-off between convenience andsafety prevents any engineering meas-ure from being automatically assumedto be successful. VRU facilities shouldalways be monitored in order to iden-tify which locations are operating suc-cessfully and where modifications areneeded to facilities not functioningproperly. Over time, a better under-standing should evolve of the factorsneeded to improve VRU safety.

ii. While improved and increased VRUfacilities are a prerequisite to VRUsafety, they cannot be expected to im-prove VRU safety on their own. En-forcement, and education and publicitycampaigns are needed to motivateproper road use behavior from allmodes. Enforcement and publicity cam-paigns should be designed and timed tocomplement and coincide with engi-neering improvement measures.

57


1) AustRoads. Traffic Engineering and Motorcycle Road Safety. Australia: AustRoads(forthcoming).

2) Bell, D. D., and C. Kuranami. 1994. “Nonmotorized Vehicles in Hanoi and Phnom Penh”:Existing Situation and Options for Improvement. Article for TRR No. 1441. Washington,DC, US: National Research Council.

3) Centre for Biomedical Engineering. 1995. The Work Done 1991-95. New Delhi, India:Indian Institute of Technology.

4) Center for Research and Contract Standardization in Civil Engineering (CROW). 1994.Sign Up for the Bike. The Netherlands: CROW.

5) Centre for Scientific and Industrial Research (CSIR). 1992. Pedestrian FacilityGuidelines: Manual to Plan, Design and Maintain Safe Pedestrian Facilities. ResearchReport RR92/126. South Africa: Department of Transport.

6) Craddock, B. S. 1993. Accidents on Pedestrian Crossings in Hong Kong. InternationalAssociation of Traffic Safety Research (IATSS), Vol.17, No. 1. Japan: IATSS.

7) Department of Transport, Bicycle Association, Cyclists’ Touring Club (CTC), andInstitution of Highways and Transportation. 1996. Cycle-Friendly Infrastructure —Guidelines for Planning and Design. UK: CTC.

8) Department of Transport. 1989. “Making Way for Cyclists.” Local Transport Note 1/89.UK: Department of Transport.

9) Department of Transport. 1995. Road Accidents Great Britain 1994: The CasualtyReport. UK: The Stationery Office.

10) DITS. 1994. Traffic Management Manual. Greater Metropolitan Dhaka IntegratedTransport Study. Dhaka, Bangladesh: UNDP.

11) Downing, A., S. Ivan, and Malik Zaheer-Ul-Islam. 1993. Pedestrian Safety in theDeveloping World. Malaysia: CARS Conference Proceedings.

12) Finch, D. J. 1995. Speed, Speed Limits and Accidents. UK: TRL.

13) Gallagher, R. 1992. The Cycle-rickshaws of Bangladesh . Bangladesh: The UniversityPress.

8 REFERENCES and BIBLIOGRAPHY

58


14) Gallagher, R. 1996. Traffic Management for Non-Motorized Vehicles and Development ofa Cycle-rickshaw Network . Draft Working Paper for World Bank. US: World Bank.

15) Goodge M. J., and T. A. Selby. 1996. Non-motorized Vehicles in the Asia-Pacific Region.Thailand: Proceedings ESCAP/ADB Road Safety Seminar.

16) Government of Bangladesh. 1993. Statistics Year Book . Bangladesh. Government ofBangladesh.

17) Government of Thailand. 1992. Road Accident Statistics. Thailand: Government ofThailand.

18) Hills, B. L., N. M. Thompson, and K. Kila. 1991. Accident Reduction Research in PapuaNew Guinea. Planning, Transportation, Research, and Computation (PTRC) 19thTransport Highways and Planning Summer Annual Meeting, Proceedings of SeminarC, 9-13 September 1991. UK: University of Sussex.

19) His Majesty’s Government of Nepal. 1996. Draft Road Safety Audit Manual. TrafficEngineering Safety Unit. Kathmandu, Nepal: Department of Roads.

20) Hoque, M., S. N. Ahmed, and A. B. Hossain. 1995. “Nonmotorized Transport: The MostDominant Mode in Bangladesh.” Paper presented at 7th World Conference on TransportResearch (WCTR). Australia.

21) Howard Humphreys and Partners Ltd. and CEBTP. 1994. Road Materials and StandardsStudy. Final Contractor Report for RHD. Bangladesh: Ministry of Communications.

22) Institution of Lighting Engineers (ILE). 1990. A Manual of Road Lighting in DevelopingCountries. UK: ILE.

23) International Conference on Traffic Safety. 1991. The Delhi Declaration on the Safety ofVulnerable Road Users. 27-30 January 1991. India.

24) JMP Consultants Ltd. 1988. Bangkok Traffic Research Project. Contractor Report 84.United Kingdom: TRL.

25) Koshi, M. 1996. Medicine Without Autopsy. CARS Conference on Asian Road Safety,Beijing, PRC, October 1996.

26) Koshi, M., ed. Leonard Evans and Richard C. Schwing. 1985. Road Safety Measures inJapan in Human Behavior and Traffic Safety. New York, US: Plenum Press.

27) Lee, S. C., and Y. I. Lee. 1996. “Safety Analysis for Conflict Behavior Between Drivers andPedestrians at Intersections.” Paper presented at the International Conference onAsian Road Safety, Beijing, PRC, 28-31 October 1996.

28) Lim, P. N. 1993. Present Status and Prospect of the Strategy of Road Traffic Safety inKorea. CARS Conference Proceedings, Kuala Lumpur, Malaysia.

29) Liu, X., L. D. Shen, and J. Huang. 1995. “Analysis of Bicycle Accidents andRecommended Countermeasures in Beijing.” Article in TRR No. 1487. US: TransportResearch Board, National Research Council.

30) Liu, X., L. D. Shen, and J. Huang. 1995. “Role of Nonmotorized Transportation in theMajor Highway System in the People’s Republic of China.” Article for TRR No. 1487. US:Transport Research Board, National Research Council.

59


31) Loss Prevention Association of India (LPAI). 1995. Road Safety Digest. Vol. 5, No. 4.India: LPAI.

32) Mathew, D. 1995. “More Bikes – Best Policy into Best Practice.” UK: Cyclists’ TouringClub.

33) Maunder, D. A. C. 1996. “Bus Accidents in the Kingdom of Nepal — Attitudes andCauses .” UK: TRL.

34) MVA. 1986. Beijing Traffic Study. PRC: ODA.

35) Mohan, P., and G. V. Soumitri. 1995. Dealing with Traffic — A Guide for Young People. India: Antara Foundation.

36) Mount, P. Safety Issues for Motorcyclists on Australian Roads in Design and MaintainingSafer Roads for Motorcycles. A summary report. Australia: Road Safety Branch,Department of Transport (unpublished).

37) National Transportation Research Center. 1990. Road Accident Analysis. PlanningCommission Report NTRC-137. Pakistan: Government of Pakistan.

38) Navin, F., and A. Bergan, J. Qi, and J. Li. 1994. “Road Safety in the People’s Republic ofChina.” TRR No. 1441. US: Transportation Research Board, National ResearchCouncil.

39) Ogden, K. W. 1996. Safer Roads: A Guide to Road Safety Engineering. UK: AveburyTechnical.

40) Ogden, K. W. 1996. “Traffic Engineering for Motorcycle Safety.” Resource PaperAustRoads Project RSG21. Australia (unpublished).

41) PADECO Co., Ltd. 1993. Nonmotorized Vehicles in Asian Cities: Part 1 — Inventory ofNeeds and Opportunities. US: Draft Final Report for the World Bank.

42) PADECO Co., Ltd. 1993. Nonmotorized Vehicles in Asian Cities: Part 1 — TechnicalGuidelines. US: Draft Final Report for the World Bank.

43) PADECO Co., Ltd. 1995. Nonmotorized Vehicles in 10 Asian Cities. US: Trends, Issuesand Policies — Report TWU 20, the World Bank.

44) Pasricha, P .S. 1983. Pedestrian Planning in Greater Bombay. Traffic Engineering andControl. London, UK: Printerhall.

45) PRC. 1995. National Road Accident Statistics. PRC: Government publications.

46) Purcell, R. H. 1981. Calcutta Traffic Engineering Project: 3 . Supporting Studies, TrafficEngineering and Control, 22 (1). UK: Printerhall.

47) Quick, A. 1991. Unequal Risks — Accidents and Social Policy. UK: Socialist HealthAssociation.

48) Ribbens, H., and S. Govert. 1997. Traffic Calming as a Means for Promoting PedestrianSafety in Urban Areas in South Africa. 76th Annual Transport Research Board Meetingin Washington, DC, US, 12-16 January 1997.

60


49) Roads and Traffic Authority and Federal Office of Road Safety. 1993. Sharing the MainStreet. Australia.

50) Ross, A. 1984. Korea Traffic Safety Study. Final Report. World Bank Provincial andCountry Roads Project. Republic of Korea: Ministry of Home Affairs.

51) Ross, A., M. Rahmatullah, and C. E. Ghee. 1996. Road Safety in the Asia-Pacific Region:Trends and Prospects. CARS Conference Proceedings, October 1996. Beijing, PRC.

52) Ross Silcock and TRL. 1991. Towards Safer Roads in Developing Countries. UnitedKingdom: TRL.

53) Royal Malaysia Police. 1994. Statistical Report on Road Accidents. Malaysia: RoyalMalaysia Police.

54) Sayer, I. A. 1990. Pedestrians in Developing Countries. International Conference on Road Safety and Accidents in Developing Countries, Egypt.

55) Sayer, I. A., and A. J. Downing. 1996. Pedestrian Accidents and Road Safety Educationin Selected Developing Countries: TRL Report 227. UK: TRL.

56) Sayer, I. A., and C. J. Baguley. 1994. Preliminary Studies of Pedestrian CrossingImprovements in Karachi, Pakistan. Project Report 68. UK: TRL.

57) Senanayake, U., and T. A. Perimpanayagam. 1993. Case for Safer Roads in Sri Lanka.Sri Lanka: Sri Lanka Police.

58) Thompson N. M. and D. Rudjito. 1991. “Traffic Accident Investigation in Indonesia.”Australia: Paper for the 7th Road Engineering Association of Asia and Australasia(REAAA) Safety Workshop, Kuala Lumpur, Malaysia.

59) Traffic Engineering and Safety Unit, Department of Roads, Nepal. 1995. Road TrafficAccidents Half-yearly Report FY 2052/2053. Nepal: Ministry of Works and Transport.

60) Traffic Engineering and Safety Unit, Department of Roads, Nepal. 1996. Draft RoadSafety Audit Manual. Nepal: Ministry of Works and Transport.

61) TRL. 1996. The Use of Traffic Signals in Developing Cities. Overseas Road Note 13.UK: TRL.

62) TRR No. 1294. 1991. “Safety and Human Performance, Nonmotorized Transportation.”US: Transportation Research Board, National Research Council.

63) TRR No. 1327. 1992. “Operations and Safety, Nonmotorized Transportation.” US:Transportation Research Board, National Research Council.

64) Yogyakarta Urban Development Project. 1995. Master Plan Transportation of Urban Agglomeration and Daerah Istimewa Yogyakarta. Draft Report. Indonesia: Direktorat Jenderal Cipta Karya.

61


9 CONTACTS

RETA PROJECT TEAM

Ross Silcock Ltd.Alan Ross, Project Director, Ross Silcock Ltd., Old Brewery Court, 156 Sandyford Road,Newcastle upon Tyne NE2 1XG, UK; Tel (work): (44 191) 261 8101; Fax: (44 191) 261 8340;Tel (home): (44 191) 265 0060; E-mail: [email protected]

Caroline Ghee, Mike Goodge, Kathleen Ness, Chris Robson, Tim Selby, and Kim Smith

TRL

Goff Jacobs , Overseas Project Director, Transport Research Laboratory, Overseas Centre, OldWokingham Road, Crowthorne, Berkshire RG45 6AU, UK. Tel: (44 1344) 773 131; Fax: (441344) 770 880

Chris Baguley and Amy Aeron-Thomas

The project team is grateful to the 112 participants from 23 countries in the Asian andPacific region who attended the ADB/UN/ESCAP seminar held in Bangkok, 2-6 September1996. The authors would particularly like to thank the chairpersons and moderators of theworkshops at the Bangkok seminar. They are as follows:

Workshop A (Engineering): Masaki Koshi (Chairperson) of Japan, R. D. Mehta ofIndia, Francis Navin of Canada, Peter Waugh of Australia, and Malik Zaheer-ul-Islam of Pakistan;

Workshop B (Enforcement): Camillus Abbeygoonewardena (Chairperson) of SriLanka, Qalo Bulatiko of Fiji, and Romeo Maganto of the Philippines;

Workshop C (Education): Milton Tan (Chairperson) of Singapore, Choong Khoon Sengof Singapore, Halimahtun Mohd. Khalid of Malaysia, Chandra Shekhar of Fiji,and M. Sadiq Swati of Pakistan; and

Workshop D (International Cooperation and Assistance): Stein Lundebye (Chairperson)of the World Bank, Goff Jacobs of TRL, and Alan Ross of Ross Silcock.

Support and advice was also received from a Steering Group comprising representatives offunding agencies, international organizations, technical groups, and individual experts from de-veloping countries. They are as follows:

John Flora of the World Bank, Alamgir Mojibul Hoque of Bangladesh, Ian Johnston ofREAAA, Charles Melhuish of the ADB, M. Ramatullah of UN/ESCAP, and C. Romerof WHO.

62


INTERNATIONAL AGENCIES

ASIAN DEVELOPMENT BANK

Charles Melhuish, Senior Policy Specialist, Asian Development Bank, P.O. Box 789, 0980Manila, Philippines. Tel: (63 2) 632 6803; Fax: (63 2) 636 2423; E-mail:[email protected]

UNITED NATIONS ECONOMIC AND SOCIAL COMMISSION FOR ASIA AND THEPACIFIC

M. Rahmatullah, Director, Transport, Communications, and Tourism Division, UN/ESCAP,The United Nations Building, Rajadamnern Avenue, Bangkok 10200, Thailand. Tel: (66 2) 2881234; Fax: (66 2) 288 1000

UN/ESCAP: Dieter Niemann, V. N. Timopheyev, I. P. Tiwari, and J. C. Witkowski

THE WORLD BANK

John Flora, Manager, Transport Division, World Bank, Room S-6029, 1818 H Street NW,Washington, DC 20433, US. Tel: (1 202) 473 8866; Fax: (1 202) 522 3223Stein Lundebye, Senior Transport Engineer, Infrastructure Operations Division, CountryDepartment 1, South Asia Region, The World Bank, 1818 H Street NW, Washington, DC20433, US. Tel: (1 202) 458 0148; Fax: (1 202) 477 5520; E-mail: [email protected]

WORLD HEALTH ORGANIZATION

C. Romer, Director, Global Reduction Programme, World Health Organization, Avenue Appia20, Ch-1211 Geneva 27, Switzerland. Tel: (41 22) 791 2111; Fax: (41 22) 791 0746

63

For more information on ADB, visit: http://www.adb.orgISBN 971-561-113-3Stock number 010499

Documents

Vulnerable Road Users_adb()