Pedestrian interaction with highly automated vehicles in shared spaces: the roles of vehicle kinematics, external human–machine interfaces, and social cues

Abstract Highly automated vehicles (HAVs; SAE Levels 4–5) will increasingly operate in mixed traffic, raising challenges for pedestrian interaction—especially in shared-space environments where coordination relies on negotiation rather than formal right-of-way rules. Prior research has largely examined one-to-one pedestrian–HAV encounters, while the role of social cues from additional pedestrians remains underexplored. This study investigated how pedestrians integrate (a) an additional pedestrian’s behavior (absent vs. present–stopping vs. present–continuing to walk), (b) vehicle kinematics of the HAV (yielding vs. non-yielding), and (c) a visual light-based external human-machine interface (eHMI; present vs. absent) when forming willingness to cross and perceived safety in a shared-space scenario. In a video-based online experiment ( N = 136; 3 × 2 × 2 within-subjects), participants rated willingness to cross and perceived safety after each 9-s rendered video clip. Vehicle kinematics had a strong influence on both willingness to cross and perceived safety. The eHMI increased willingness to cross and perceived safety under yielding kinematics, but had no effect under non-yielding kinematics, indicating that eHMIs are beneficial primarily when congruent with vehicle kinematics. Additional pedestrian behavior affected both willingness to cross and perceived safety, but only when the HAV displayed yielding kinematics. Together, the findings suggest that, in shared-space interactions, vehicle kinematics emerged as the dominant cue, while eHMIs and social cues act as supportive signals whose influence depends on the vehicle’s kinematic. Further research is recommended to explore the influence of multiple pedestrians and more complex traffic scenarios.