Understanding traffic law enforcement perceptions of autonomous vehicles: Insights from Chinese cities with operational robotaxi fleets through hypothesis testing and machine learning

Autonomous vehicles (AVs) are expected to improve road safety by reducing human error, which accounts for over 90% of traffic incidents. However, AVs also present new challenges for traffic law enforcement, particularly in their interactions with officers and their ability to navigate emergency situations, such as safely passing law enforcement officers (LEOs). This study examines these challenges through a survey of 1522 traffic LEOs in Chinese cities with operational AV fleets. The survey focused on two critical perceptions: LEOs’ confidence in AVs’ ability to follow traffic commands and trust in AVs’ ability to safely pass LEOs. Results indicate that over 90% of LEOs lacked confidence in AVs’ ability to follow commands, and nearly 70% did not trust AVs to safely pass LEOs. Officers with direct experience managing AV-related incidents or riding in a robotaxi reported slightly higher confidence and trust levels, but overall perceptions were mixed. Chi-square tests and machine learning analyses further indicated that factors such as AV-related knowledge, procedural training, and attitudes toward robotaxis were significantly associated with these perceptions. These findings underscore the need for targeted training programs, increased exposure to AV technologies, and public engagement strategies to enhance law enforcement trust in AVs and their safe integration into traffic systems.