Analyzing Frontage Roadway Crashes through the Lens of Probabilistic Graphical Models

Frontage roads play a crucial role in providing access and connectivity between freeways, principal arterial roadways, and adjacent commercial and non-commercial areas in the U.S. In Texas, frontage roads are a key design solution to facilitate access along rural freeways and principal arterial routes. These roads typically function as two-way roads in less developed urban and rural regions, while in urban and city-centered areas, they predominantly operate as one-way roadways. Despite their importance in enhancing accessibility, frontage roads also pose significant safety concerns, which have not been thoroughly investigated until now. To address this knowledge gap, this study examined safety issues related to frontage roads by analyzing six years (2014-2019) of frontage road crash data (235,522 crashes) collected in Texas. The primary objective is to develop conditional probability models that shed light on the causal patterns of crash occurrences on these roadways. The study identified key crash characteristics, including road design elements and environmental conditions, shedding light on high-risk scenarios such as intersections, divided roads, and specific weather and lighting conditions, thereby providing valuable insights for targeted safety interventions and policy decision. The results can provide valuable evidence for policymakers, traffic engineers, and transportation agencies to develop evidence-based safety strategies and policies. The findings will be instrumental in guiding safety improvements and traffic engineering interventions tailored to frontage roads, ensuring safer travel experiences for road users, including pedestrians and cyclists, and reducing the potential for traffic conflicts and collisions.