Crash Severity and Rate Evaluation of Conventional Vehicles in Mixed Fleets with Connected and Automated Vehicles

Abstract

Abstract Connected and Automated Vehicle (CAV) technology, although in the development stage, is quickly expanding its market but a full market penetration might not be rapid. The safety concern is the paramount challenge to widespread adoption of this disruptive technology. During the transition period, fleets will be composed of a combination of CAVs and conventional vehicles, and therefore it is germane to investigate the repercussions of CAVs on traffic safety at different penetration. Since crash severity and frequency in conjunction reflect the traffic safety, this study attempts to investigate the effect of CAVs on both crash severity and frequency. PTV VISSM microsimulation platform is used to simulate M1 Geelong Ring Road network (Princes Freeway) in Victoria, Australia, which is the testbed for this study. Network performance is evaluated using performance metrics (Total System Travel Time, Delay and instantaneous speed profiles). Surrogate safety measures (time to collision, post encroachment time, etc.) are examined to inspect the safety in the network. The results showed that CAVs would not inevitably decrease the crash severity and crash rate involving manual vehicles, despite the improvement in network performance, given the demand and the set of parameters used in our operational CAV algorithm are intact. Additionally, the study identifies that the safety benefits of CAVs are not proportional to CAV penetration, and a full-scale benefits CAVs can only be achieved at 100% CAV penetration. The results presented in this study provide an insight into the repercussion of CAVs on comprehensive traffic safety to the insurance companies and other industry participants, enabling safety-related services and more enterprising business models.