Research Papers

Spatial Distribution of Pedestrian-Motor Vehicle Collisions Before and After Pedestrian Countdown Signal Installation in Toronto, Canada

Version 1
Date added June 27, 2017
Downloaded 0 times/fois
Category 2017 CARSP XXVII Toronto
Tags Research and Evaluation, Session 4B
Author/Auteur Linda Rothman
Stream/Volet Research and Evaluation

Slidedeck Presentation Only (no paper submitted)



Pedestrian countdown signals (PCS) have been installed in many cities over the last 15 years. The few studies that have examined the effects of PCS on actual pedestrian motor vehicle collisions (PMVC) have had varied results. An important contextual factor often not considered is the spatial distribution of collisions related to PCS. Examining the spatial distributions of PMVCs can provide some indication of the effectiveness of PCS in the context of differing road environments, and for populations known to be more vulnerable, such as children and older adults. To compare the spatial patterns of collisions pre and post installation of PCS at intersections with PCS and roadways and intersections without PCS in the City of Toronto, and examine differences by age. Countdown signals were installed at the majority of Toronto signalized intersections from 2007 to 2009 (1864 intersections). Spatial patterns of both PCS and non-PCS police-reported collision sites were examined four years prior to and after the installation of the signals. The spatial distribution of collisions were estimated using kernel density estimates for all collisions and stratified by age before and after installation. Raster count maps were created over the area to examine the difference in the number of collisions before and after installation. Areas of higher or lower point density were identified pre to post installation. There were 15,164 collisions included. Some PCS locations had more collisions after PCS installation; particularly for adult and older adult pedestrians. Increased adult pedestrian collisions were concentrated downtown, whereas older adult collision increases occurred throughout the city following no spatial pattern. There was a reduction in collisions involving children at both PCS and non-PCS locations, with greater reductions at non-PCS locations. There was generally a more consistent reduction in collisions post installation at the non-PCS locations compared to the PCS locations. PCS can improve pedestrian safety; however, our descriptive spatial pattern results illustrate that their potential effect on road safety varies by age and location. The overall reduction in collisions involving children at both PCS and non-PCS locations could be an indicator of the reduction in walking mode share in children over the study period in Toronto, as decreases in walking in children have been well documented. PCS effectiveness may also be related to concomitant changes in the roadway built environment. Areas where post-PCS installation indicated increased collisions could be examined for changes in signal timing, roadway design or vehicle turning restrictions that could yield desired safety benefits. PCS can improve pedestrian safety, but their installation may have the unintended effect of increasing PMVC under certain circumstances and at specific locations. Knowing that the effect of PCS varies within a city suggests that their effect will be different between cities and can help explain discrepancies in the literature. The age and location effects of PCS need to be fully understood and accounted for to consistently improve pedestrian safety.

Linda Rothman