Predicting pedestrian crash locations in urban India: An integrated GIS-based spatiotemporal HSID technique

Abstract

Abstract Pedestrians are one of the most vulnerable road users globally. Recent years have witnessed an increasing interest among the scientific community to analyze and enhance pedestrians' safety in an environment dominated by motor vehicles. This study proposes a three-step methodology to identify current and future critical pedestrian crash hotspots. Firstly, available multi-year crash data from two cities in India is digitized, and the spatial autocorrelation tool is used to determine the pedestrian crash hotspots. Secondly, space-time cube and emerging hotspot analysis are carried out to predict crash hotspots along urban streets. Finally, Hotspot Identification (HSID) methods, i.e., Equivalent Property Damage Only (EPDO) and Upper-tail Critical Tests are used to rank the road links based on spatio-temporal crash severity leading to the identification of links needing urgent interventions. The proposed three-step integrated methodology is novel and has never been used to simultaneously identify and prioritize the critical pedestrian crash locations as it has been done in the present study. The developed methodology identifies sections of arterial roads—Strand Road and AJC Bose Road in Kolkata and Gota Road in Ahmedabad, as the critical hotspot links that require urgent intervention.