Assessment of connected and autonomous vehicles impacts on urban road safety through microsimulation

Abstract

The integration of Connected and Automated Vehicles (CAVs) into urban environments promises significant improvements in road safety and traffic management, particularly at full market penetration. However, the transitional phase of CAV adoption remains poorly understood due to the limited and localized presence of these vehicles. This study aims to evaluate the impact of CAVs on road safety in an urban corridor by examining their longitudinal behavior across varying levels of market penetration and traffic demands. Utilizing the VISSIM microsimulator, this research implemented the Cooperative Adaptive Cruise Control algorithm, designed for longitudinal control of CAVs and incorporating vehicle-to-infrastructure communication. Given that microsimulation does not directly model crashes, the study applied safety continuum theory to assess conflicts—potential crashes arising from risky vehicle interactions. The analysis focused on rear-end conflicts, examining both frequency and severity. Results showed that during the initial transitional period, CAVs could increase conflict occurrences—up to six times in low vehicular demand and twice in high demand—primarily involving conventional vehicles as followers. However, scenarios with full CAV automation demonstrated a 54% reduction in conflicts compared to the base scenario, indicating CAVs potential to improve road safety. Additionally, the study employed a multinomial logit model to assess the impact of CAVs on conflict severity, finding that the probability of high-severity conflicts decreased with greater CAV penetration. This suggests that while conflict frequency may rise during initial deployment stages, the severity of these conflicts does not escalate correspondingly, emphasizing the potential benefits of CAVs in reducing severe accidents.