A Surrogate Safety Analysis at Protected Freeway Ramps Using Cross- Sectional and Before-After Video Data

摘要

This study presents a surrogate approach for safety analysis of freeway facilities using automatedtrajectory collection and behavioural analysis from surrogate measures of safety (in particular time tocollision). This methodology is proposed as a potential alternative or complement to the classicalapproach based on historical accident data, particularly suited for evaluating the microscopic safetyeffects of road treatments for which there is a lack of traffic and accident data. A short theoreticaldiscussion of traffic conflicts is followed by a proposed methodology illustrated using as a small sampleof freeway ramps as an application environment. From this sample, video data is obtained as part of asafety study to investigate the effectiveness of the "one-way lane-change ban" treatment near urbanfreeway ramps in Montreal, Canada. To illustrate the applicability of our methodology, two comparativeexamples are presented: (1) a cross-sectional study and (2) a before-after study involving two sites, one ofwhich had video data available before and after the implementation of the treatment. Various methods ofaggregating the data, spatially and temporally, are explored in the applications.