Can Connected Autonomous Vehicles Improve Mixed Traffic Safety Without Compromising Efficiency in Realistic Scenarios?

CAVs have the potential to improve traffic safety and efficiency in a fully connected environment, but they face additional challenges when they interact with human-driven vehicles in mixed-traffic scenarios due to the uncertainty in humans’ driving behaviour. Another challenge is the unreliability of communication networks. Delays and packet losses in communication links due to dense traffic, communication interference, channel fading, etc. may jeopardize CAVs driving behaviour and negatively affect traffic safety and efficiency. While the impact of CAVs on traffic safety and efficiency at different market penetration rates (MPRs) has been studied extensively, CAVs have been assumed to operate in either realistic traffic scenarios with perfect communication or with only a very small number of vehicles with imperfect communication. In contrast, this paper investigates the impact of CAVs on both traffic safety and efficiency in realistic scenarios in terms of imperfect communication, large-reaction time, vehicle modelling, and traffic scenario, with a focus on quantifying the detrimental effects of unreliable V2V communication and large-reaction time. Results show that CAVs can significantly improve mixed traffic safety in the presence of unreliable V2V communication and large-reaction times by adopting a very cautious car-following strategy (i.e., larger time headways), but this is at the expense of a slight reduction in traffic efficiency. Specifically, safety conflicts can be reduced by up to 66.67% at 70% penetration rate in the presence of packet drop of 70% and reaction time of 0.5s, but only at the cost of a travel time increase of 2.6%.