Title

Safety Performance Functions for Signalized Intersections in Large Urban Areas: Development and Application to Evaluation of Left-Turn Priority Treatment

Document Type

Journal Article

Publication Date

2005

Subject Area

economics - appraisal/evaluation, infrastructure - bus/tram priority, infrastructure - traffic signals, mode - bus, operations - performance, operations - traffic, organisation - performance, place - urban, planning - safety/accidents

Keywords

Urban areas, Traffic signal priority systems, Traffic signal preemption, Toronto (Canada), Signalized intersections, Signalised intersections, Sideways collisions, Sideswiping accidents, Side impact collisions, Side collisions, Safety performance functions, Preemption (Traffic signals), Left turns, Empirical Bayes method, Bus priority

Abstract

This paper describes the development of safety performance functions (SPFs) for 1,950 urban signalized intersections on the basis of 5 years of collision data in Toronto, Ontario, Canada. Because Toronto has one of the largest known, readily accessible, urban signalized intersection databases, it was possible to develop reliable, widely applicable SPFs for different intersection classifications, collision severities, and impact types. Such a comprehensive set of SPFs is not available for urban signalized intersections from data for a single jurisdiction, despite the considerable recent interest in use of these functions for analyses related to network screening, and the development, prioritization, and evaluation of treatments. The application of a straightforward recalibration process requiring relatively little data means that the SPFs calibrated can be used by researchers and practitioners for other jurisdictions for which these functions do not exist and are unlikely to exist for some time. The value of the functions is illustrated in an application to evaluate a topical safety measure�”left-turn priority treatment for which existing knowledge is on a shaky foundation. The results of this empirical Bayes evaluation show that this treatment is quite effective for reducing collisions, particularly those involving left-turn side impacts.