A Formal Approach to Road Safety Assessment Using Traffic Conflict Techniques

Abstract

Traffic conflict techniques enable a comprehensive assessment of traffic safety analysis. Formal methods allow the identification of factors that contribute to traffic safety issues and provide evidence of potential safety degradation. As such, formal methods provide a novel way to model traffic rules and verify road users' compliance. The paper proposes formalizing a traffic safety rule in differential dynamic logic and using KeYmaera theorem prover for verification. This rule considers time-to-collision (TTC), space headway (SHW), and shockwave speed (SWV). To validate the effectiveness of this rule in realistic traffic scenarios, we conducted a study using calibrated microsimulation data from the SR528 highway in Orlando, Florida. Our analysis examined the TTC, SHW, and SWV values for vehicle platoons on the highway and demonstrated how smaller TTC and SHW values indicate shockwaves and subsequent conflicts. Furthermore, we observed that shockwave speed could contribute to traffic conflicts by enabling evasive actions such as sudden braking or lane changes as the risk of collisions increases. By highlighting these findings, we aim to provide valuable insights into the real-world applicability of formal methods for traffic safety and their potential in promoting safer driving practices that can help create reliable autonomous vehicle control systems.