Assessing the Impacts of Transit Signal Priority on Crash Severity: An Empirical Assessment Using Bayesian Logit Model with Unobserved Heterogeneity

Abstract

Transit signal priority (TSP) is a traffic management strategy to enhance the quality of public transit service while also providing substantial safety benefits. This study used a binary Bayesian logit model with random effects, which accounts for unobserved heterogeneity, to explore impacts of TSP on severity of corridor-related crashes in Florida. The analysis revealed deploying TSP was linked to lower crash severity, reducing the likelihood of fatal plus injury (FI) crash by 7.96%. The study also investigated other factors contributing to crash severity, including crash characteristics, driver characteristics, roadway geometry, and environmental factors. Distracted driving, vulnerable road users, and higher speed limits increase the risk of FI crashes, while higher average annual daily traffic (AADT) was linked to lower risk. Certain crash types, such as rear-end and sideswipe crashes, were also associated with lower risk of FI crashes. These findings hold crucial implications for transportation agencies when planning future TSP deployments.