Balancing safety and efficiency for autonomous vehicles at urban uncontrolled crosswalk: challenges and countermeasures

Abstract

At uncontrolled crosswalks, defensive autonomous vehicles (DAVs) prioritize yielding to pedestrians but may exacerbate congestion. Interactive autonomous vehicles (IAVs) attempt to prioritize pedestrian passage under safe conditions; however, they demand more advanced technology, and their practical application remains uncertain. Platoon control presents a potential solution to mitigate the limitations of both driving styles. This study develops an agent-based pedestrian-vehicle interaction framework and introduces a platoon formation module that accounts for passenger comfort. The study examines the impact of combining driving styles (DAV and IAV) and the platoon control method on the safety and efficiency of pedestrian-AV interactions. Results indicate that compared to DAV, IAV increases interaction events by 32.2%, yet achieves a yielding compliance rate of only 40% while reducing total delays by 60.7%. Platoon control effectively enhances the safety and efficiency of both driving styles, with its benefits becoming more pronounced as pedestrian and vehicle traffic volumes increase. This study provides methodologies and strategies to address the challenges of integrating AVs into urban environments.