View
45
Download
3
Category
Preview:
Citation preview
CHANGING BEHAVIOURS
THROUGH USER INTERACTION
PRESENTED BY URIE BEZUIDENHOUT
DA VINCI TRANSPORT PLANNING
UNIVERSITY OF AUCKLAND
SO HOW BAD IS CONGESTION?
It depends on how you define congestion and what you compare it to
$1billion per a?
Auckla nd Wel l ing t on Chr ist church
$702
m
$101
m
$77
m
$54
0
$21
0
$20
4
Congestion cost (2004)Total ($mpa) Person $pa)
WHAT CAN WE DO ABOUT IT?
Now
Autonomous vehicles
Medium to Long termShort Term
Medium Term
INFLUENCING DRIVER BEHAVIOUR
Now
Short Term
TRAJECTORY PLOTS
FLUSHING OUT QUEUES
Queued @ Red
QueuedPlatoon
Detector Zone
Free flow Platoon
S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S270
1
2
3
4
5
6
7
8
9
10
11
NEW Q-Predict Model Accuracy
Cycle Reference
Q -L
engt
h (v
eh)
RESEARCH CASE STUDY – Auckland CBD
EXPERIMENTAL EVIDENCESense traffic
Optimise signals
Drivers divert
Route flow
changes
Signal timing
changes
OBJECTIVES
• Quantify theoretical maximum efficiency
• Signals can estimate queue
• Optimise OFFSET on cycle-by-cycle – queue minimisation
• Only selected corridors optimised
• Apply MUSIC - OPERA algorithm
• Drivers alter routes in response to signal changes
RESULTS – NETWORK WIDE
Short term (Medium term)• Cyclical queues - 8% (-
28%)• Overcapacity queues - 4% (-31%)• Stop-start queues - 39% (-59%)• Travel time reduction - 5% (-20%)
OPERA OPTIMISING BENEFITS
-25 % travel time
$ 400 Savings/person/year
$0.25 billion /yearTravel time savings for Auckland
500,000 trips in peak hr
11 min trip in peak hr
• Competing signage
• Short block lengths
• Visually complex
• Queues
• Pedestrians and cyclists
DRIVER BEHAVIOUR
13
DRIVER WORKLOAD
+ 6 s ~ 90 m
14
TASK ANALYSIS - INTERSECTIONS
15
FIXATION
GAZE HEATMAP
DYNAMIC HAZARDS
DYNAMIC HAZARDS
CALIBRATING DRIVERS
• Use queues to discourage movements• Flush out queues before arrival• Greenwave along alternative route• 4 - 6 weeks to get measurable benefits
Recommended