Sekadakis M., Oikonomou M., Ziakopoulos A., Yannis G., “Predicting Conflict Severity and Safety Impacts of Automated Vehicles in Simulated Urban Networks”, Proceedings of the Transport Research Arena TRA 2026 Conference, Budapest, 18-21 May 2026.

This study investigates how conflict severity and Time-To-Collision (TTC) in urban traffic environments are affected by vehicle dynamics and the introduction of automated shuttle services. Microscopic simulation data from five European pilot sites were analysed using two statistical models. A binomial logistic regression was employed to classify conflicts as high or low severity, considering variables such as the maximum change in velocity (MaxDeltaV), the maximum deceleration of the second vehicle, conflict type (rear-end, lane change, or crossing), and the presence of automated vehicles. The results indicate that higher MaxDeltaV increases the probability of severe conflicts, while greater deceleration reduces it. Importantly, automated vehicle deployment is associated with a statistically significant reduction in severity. In parallel, a loglinear regression model was applied to TTC, showing that larger velocity changes shorten TTC, whereas stronger braking and the introduction of AVs extend it. Overall, the findings suggest that automated shuttles enhance traffic safety by lowering the likelihood of severe conflicts and providing additional time for evasive action across the network.