Traffic%20Generation%20for%20Studies%20of%20Gap%20Acceptance

1
Traffic Generation for Studies of Gap Acceptance

• Joseph Kearney
• Timofey Grechkin
• James Cremer
• Department of Computer Science
• Jodie Plumert
• Department of Psychology
• University of Iowa

2
Bicycling Injuries

• Bicycle crashes are a common cause of severe
  injury in childhood
• ERs treat 500,000 bicycle-related injuries a year
• Highest injury rate is among children 5-15
• Motor Vehicles involved in 90 of bicycling
  fatalities
• Prevention
• Need to understand why car-bicycle collisions
  occur
• How do immature cognitive and perceptual skills
  put children at risk for car-bicycle collisions?

3
Gap Acceptance in the Hank Bicycle Simulator
4
Bicycle Simulator Video
5
Research Questions

• Are there age differences gaps selection?
• How are gap choices related to crossing behavior?
• Do gap choices change in dense traffic?
• How is temperament related to risk in road
  crossing?

6
Prior Work on Gap Acceptance

• Critical Gap Estimation
• Important component of flow computations
• Focus on gap selection
• Crossing behavior not examined
• Gap Affordance
• How is gap selection related time to cross?
• Converging evidence that people do not account
  for diminished skill
• Child pedestrians
• Alcohol impaired pedestrians
• Elderly with attentional deficits
• Child bicyclists

7
Virtual Environment

• Three 10X8 ft screens (rear projection)
• Projection Design Projectors -1280x1024
  pixels/screen
• Square (Cave-like) configuration
• Seven networked PCs
• Dynamic pedal torque

8
Road Crossing Experiment(Plumert, Kearney,
Cremer, Child Development 7(4), 2004)

• Subjects Sixty 10- and 12 year olds and adults
• Procedure
• Warmup 3 blocks with no traffic
• Gap crossing
• 6 intersections with steady traffic at 25 or 35
  MPH
• Random gaps (1.5, 2.0, 2.5, 3.0, 3.5, 4.0
  seconds)
• Subjects instructed to stop at each intersection
  and safely cross
• Measures
• Gap choice
• Time to spare
• Wait time
• Stopping

9
Coding The Data Visualizer
10
Summary of Road Crossing Results

• Children and adults chose the same size gaps
• Average size of 3.5-3.6 seconds
• Children had less time to spare
• when cleared path of car, on average
• 10-year olds had 1.13 sec to spare
• 12-year olds had 1.49 sec to spare
• Adults had 1.98 sec to spare
• Why did children have less time to spare?
• Started later
• Took more time to get going

11
Long Wait Experiment

• Subjects 120 10- and 12 year olds and adults
• Procedure
• Warmup
• Gap crossing
• 4 intersections with random gaps
• 4 long wait intersections
• 4 intersections with random gaps
• Long Wait Traffic
• 8-10 uncrossable gaps (1.5 and 2 s)
• Stair-step increase in gap size
• Alternating (two crossable gaps four
  uncrossable gaps)
• 1.5, 2.0, 1.5, 1.5, 2.0, 1.5, 1.5, 2.0, 3.0,
  3.0, 1.5, 2.0, 2.0, 1.5, 3.5, 3.5, 1.5, 1.5, 2.0,
  1.5, 4.0, 4.0,

12
Long Wait Gap Choice
1.0
0.9
Middle four
intersections
Last four intersections
0.8
0.7
Gaps Taken
0.6
Gaps Seen
0.5
0.4
first four intersections
0.3
0.2
0.1
5.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Gap Size
13
Summary of Long Wait Results

• Children and adults chose the same size gaps
• Children had less time to spare when they cleared
  path of oncoming car
• Both children and adults accept much smaller gaps
  at intersections with dense traffic
• This risky behavior carries over to intersections
  with normal traffic density

14
Analysis of Gap Acceptance Data

• Average gap selected is biased
• Over-estimates critical gap
• Cautious drivers are over-represented
• (Gattis and Low, 1999)
• Logistic Regression
• Estimates critical gap
• Cautious drivers still over-represented
• Stair-step presentation
• Long waits may influence response criteria
• Savvy drivers may wait

15
How Do Children and Adults Cross Mutli-Lane
Traffic?

• Requires passage through two overlapping gaps
• More difficult perceptual task spatially and
  temporally
• Greater payoff for anticipation
• Greater overall distance to be crossed
• Staged crossing through rolling gaps
• Improving Pedestrian Safety at Unsignalized
  Crossings
• TCRP Report 112, 2006

16
Gap Acceptance with One-Lane Traffic
Tail
Lead
17
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
18
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Near lane
time
current time
19
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Near gap
Near lane
time
current time
20
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far gap
Far lane
Near lane
time
current time
21
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Overlap
Near lane
time
current time
22
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Near lane
crossing interval
time
current time
23
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Near lane
crossing interval
time
current time
24
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Near lane
crossing interval
time
current time
25
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Near lane
crossing interval
time
current time
26
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Near lane
crossing interval
time
current time
27
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Near lane
crossing interval
time
current time
28
Gap Acceptance with Two-way Traffic
Tail far lane
Lead far lane
Lead near lane
Tail near lane
Far lane
Near lane
crossing interval
time
current time
29
Generation of Two-way Traffic

• Specify Between Lane Gaps
• Randomly generate gaps
• (time between successive arrivals irrespective of
  lane)
• Randomly assign lane
• Produces natural clusters and breaks
• May reduce problem to one-lane crossing
• Specify Within Lane Gaps
• Independent (randomized) streams on two lanes
• Produces steady stream of two-way traffic
• Requires synchronization of gaps
• Increased gap size

30
Future Work Young Drivers

• How are young drivers gap choices related to
  their crossing behavior?
• Little is known about road-crossing behavior in
  young drivers
• Studies of road-crossing require a wide FOV
• Scenarios apply to driving

31
Summary

• Gap Crossing
• Essential skill for safe driving
• Complex perceptual task
• Detect temporal relations in spatially disparite
  dynamic streams
• Complex motor task
• Synchronize movement to multiple temporal
  intervals
• Anticipate arrival times
• Probe for investigating skill of driving
  populations
• Critical events vs. prosaic driving tasks
• Understand perceptual and motor skills needed for
  safe driving
• Investigate differences in driving populations
• Identify risky behaviors
• Improve Training

32
Acknowledgments

• NSF Support INT-9724746, EIA-0130864, and
  IIS-0002535 National Center for Injury
  Prevention and Control R49/CCR721682
• National Institutes of Health 1 R01
  HD052875-01
• Contributing students, staff, faculty
• Hongling Wang Geb Thomas
• David Schwebel Pete Willemsen
• Penney Nichols-Whitehead Scott Davis
• Jennifer Lee Steffan Munteanu
• Sarah Rains Joan Severson
• Sara Koschmeder Tom Drewes
• Ben Fraga Forrest Meggers
• Kim Schroeder Paul Debbins
• Stephanie Dawes Bohong Zhang
• Lloyd Frei Zhi-hong Wang
• Keith Miller
  Xiao-Qian Jiang
• Timofey Grechkin Christine Ziemer