Understanding Driver Contributions to Rear-End Crashes on Congested Freeways and their Implications for Future Safety Measures

Abstract

Empirical evidence for occurrences of shockwaves as a primary cause of rear-end crashes on freeways is well documented in the transportation safety literature. However, existing studies fail to provide a satisfactory explanation of why some shockwaves produce rear-end crashes and others do not. In pursuit of answering such a question, my doctoral research focuses on understanding the behavior of individual drivers involved in brake-to-stop events on congested freeways, and using this understanding to evaluate the implications for future safety measures. Using video recordings of shockwaves from a section of a congested freeway my doctoral research verifies a sufficient condition for a rear-end collision to happen when a sequence of drivers interact with each other in a brake-to-stop situation. Then drawing on classic results from the theory of random walks, it is possible to estimate the probability that successive braking by a platoon of drivers results in a rear-end crash. Finally, as the ultimate goal of this research is to understand the underlying mechanism that governs the behavior of drivers involved in rear-end events, this research treats drivers in a brake-to-stop event as engaged in strategic interactions in the roles of a leader and follower, who aim to maximize individual utilities. This leads to a population game whose steady state distribution describes the long-run behavior of the drivers in the population. The proposed framework is then extended to investigate (a) the safety implications of mixtures of human-operated and automated vehicles (b) the safety implications of negligence-based liability policies where individual drivers are penalized based on degree of causal contribution to the crash.