Evaluation of Pedestrian Countdown Signals

1
Evaluation of Pedestrian Countdown Signals

• Scott Washburn, Asst. Prof.Deborah Leistner,
  Grad. Research Asst. Byungkon Ko, Grad. Research
  Asst.

2
Introduction

• Flashing DONT WALK often misunderstood.
• Countdown signal provides ped with additional
  informationtime until solid DONT WALK

3
Introduction

• Hypothesized that peds will make better decisions
  with countdown information
• Intended result is that smaller of peds will be
  in x-walk at start of DW period
• Also purported that compliance with ped signal
  indications will improve
• However, it is possible that countdown
  information will lead to higher of peds
  entering x-walk on FDW

4
Research Approach

• Before-and-After Study

5
Study Sites

• W. University Ave / 20th St
• Next to campus, moderate ped activity
• W. University Ave / 17th St
• Next to campus, very high ped activity
• W. University Ave / 2nd St
• Downtown (nightclub district), very high
  weekend night ped activity
• E. University Ave / 1st St
• Downtown (professional district), moderate ped
  activity
• Archer Blvd. / SW 16th St
• Next to hospital, high ped activity

6
Site Locations
7
W. University Ave / 20th St
8
W. University Ave / 17th St
9
W. University Ave / 2nd St
10
E. University Ave / 1st St
11
Archer Blvd. / SW 16th St
12
Data Collection Periods

• Before data collection
• 9/30/03 11/1/03
• Countdown signals installed
• 10/28/03 11/04/03
• After data collection
• 11/17/03 12/13/03
• 3/24/04 4/15/04

13
Data Collection Method
14
Data Reduction

• Recorded Events
• time of pedestrian arrival at the curb
• whether pedestrian push button (if present) was
  pressed
• what phase of pedestrian signal the pedestrian
  entered the crosswalk (i.e., WALK, FDW, DW), and
  during which signal cycle (e.g., current cycle or
  a following cycle)
• phase during which pedestrian exited the
  crosswalk
• any erratic crossing behavior (e.g., running,
  turning around)
• any pedestrian-vehicle conflicts

15
Data Reduction

• Performance Measures
• of pedestrians entering the crosswalk during
  the WALK, FDW, and DW indications,
• of pedestrians exiting the crosswalk during the
  WALK, FDW, and DW indications
• compliance with the WALK indication
• of pedestrians hesitating, running, or going
  back to the curb
• of pedestrian-vehicle conflicts

16
Data Reduction

• Questions Addressed
• Are pedestrians more or less to comply with the
  WALK indication?
• Are pedestrians more or less likely to enter the
  intersection during the FDW phase?
• Are pedestrians more or less likely to still be
  making a crossing movement when the DW phase
  appears?
• Is erratic pedestrian crossing behavior more or
  less likely?,
• Are pedestrian-vehicle conflicts more or less
  likely?

17
Data Reduction

• From 8 to 24 hours of data for each intersection
  for both before and after periods

18
Analysis

• Test for significant differences in before and
  after population proportions for each performance
  measure
• Test statistic is calculated by,

19
Analysis
20
Results

• W. University Ave / 17th St
• Proportion of peds entering on WALK increased
• Proportion of peds entering on DW decreased
• Proportion of peds exiting on FDW increased
• Proportion of peds exiting on DW decreased
• W. University Ave / 20th St
• Same as for 17th
• Additionally, running increased
• Archer Blvd. / SW 16th St
• Same as for 17th
• Additionally, running decreased

21
Results

• W. University Ave / 2nd St
• Proportion of peds exiting on FDW increased
• Proportion of peds arriving on FDW and waiting
  for ensuing WALK increased
• Running increased
• E. University Ave / 1st St
• Proportion of peds entering on WALK decreased
• Proportion of peds entering on FDW decreased
• Proportion of peds entering on DW increased
• Proportion of peds exiting on FDW decreased
• Proportion of peds exiting on DW increased
• Running increased

22
Summary

• For 4 of the 5 intersections
• The countdown signals appear to have had an
  overall positive influence on crossing behavior
• Proportions of pedestrians entering on the WALK
  indication increased, while proportions entering
  on the solid DONT WALK indication decreased.
• The countdown signals have also had the effect of
  increasing the proportion of pedestrians exiting
  on the FDW interval as opposed to the DW
  interval.

23
Summary (cont.)

• The countdown signals have not had the
  potentially negative effect of increasing the
  proportion of pedestrians entering the crosswalk
  during the FDW interval,
• For three of the intersections, the countdown
  signals had the effect of increasing the
  proportion of pedestrians running in the
  crosswalk. For one intersection, this proportion
  decreased.

24
Questions ?