Identifying Corridor-Level Safety Improvements for Urban and Suburban Arterials in Florida Within a Safe System Framework

Abstract

Many agencies have adopted a Safe System approach to improving roadway safety. The Highway Safety Manual (HSM) provides methods for assessing safety, but these models are site-specific and require extensive data, making them difficult to use at large scales. This paper develops a corridor-level methodology for holistically looking at corridors made of consecutive intersections and roadway segments to identify safety improvements which align with the Safe System approach while requiring less data than HSM methods. Using a standardized definition, a total of 549 corridors on urban and suburban arterials across Florida were identified which experienced over 10,000 fatal and serious injury (FSI) crashes from 2017 through 2021. A negative binomial regression model was developed to predict mean FSI (MFSI) crashes at the corridor level (using corridor length as exposure), with the predicted values adjusted using the Empirical Bayes method to provide more accurate results. The significant factors in the model were traffic volume, intersection densities and sizing, area type, bus stop presence, citation rate, and corridor lighting presence. Increasing citation rates (citations/year/mile) for unsafe driving behaviors by one unit was predicted to reduce MFSI crash frequency in corridors by 2%, and corridors without lighting were predicted to experience 2.79 times more MFSI crashes compared with corridors with lighting. Two sister corridors in South Florida with similar roadway characteristics but different crash frequencies were also analyzed. Improvements to lighting and access control in the identified high-risk corridor could help reduce FSI crashes. Overall, this corridor approach can help agencies proactively improve roadway safety.