Using Surrogate Measures to Evaluate the Safety of Autonomous Vehicles
Youssef El-Hansali, Siham Farrag, Ansar-Ul-Haque Yasar, Elhadi Shakshuki, Khalid Al-Abri
Abstract
Automated vehicle (AV) technologies are fast, new, and have been introduced to create transport systems that enhance road safety and mobility. However, evaluating their safety has been difficult due to a lack of AV exposure data. This paper investigates the safety impact of AVs using a simulation-based surrogate safety-measure approach. Three measures of safety were chosen: time-to-collision (TTC) distribution, post-encroachment time (PET), and conflict points. The conflict-point output was also used to estimate the number of crashes per year. This paper’s results could provide useful insights for network operators about traffic conditions and road safety after the implementation of AVs. The results show that AVs improve safety; the number of crashes was reduced by about 12%.