Systemic Predictive Safety Analysis of Pedestrian Crashes for Montgomery County’s Vision Zero Program

Abstract

The goal of Vision Zero is the prevention of all traffic fatalities and serious injuries. Although traditional transportation planning is reactive to locations where serious crashes occur, some agencies are taking a more proactive approach to safety to improve locations with high expected crashes before someone is seriously injured or killed. This paper presents the results of a systemic safety analysis that produced two pedestrian-related safety performance functions for Montgomery County, MD, including 1) motor vehicle crashes with pedestrians at intersections at night and 2) through-movement motor vehicle crashes with pedestrians at road sections. These models were built using negative binomial regression of police-reported crash data collected from 2015 to 2019 for most of the county road network integrated with land use-, demographic-, and roadway variables collected by the Montgomery County Planning Department for 16,387 intersections (stop-controlled and signalized) and 29,715 segments (all functional classifications except freeways). Both models identified key transportation-related exposure variables, including motor vehicle and pedestrian volumes, proximity to transit, and crosswalk locations; they also presented land use contexts that may explain where pedestrians are likely to walk and be exposed to crash risks. These results build on current systemic safety literature and demonstrate the data collection and analysis methods that can be used in a county-level Vision Zero context to improve safety for all who walk. This paper summarizes the analysis approach, including exposure modeling, crash modeling, and applications for identifying both high-risk locations and potential mitigations. Considerations for equity and long-term planning are also discussed.