Coordinating Internal and External HMIs for Cooperative Driver-Pedestrian Situation Resolution

Traffic interactions between vehicles and pedestrians are inherently uncertain due to the dynamic nature of the interaction partners’ behavior. Resolutions require mutual understanding of the situation and each other’s intention, typically achieved through explicit and implicit communication. Assistance systems with internal or external HMIs can reduce uncertainty for drivers or pedestrians, but joint resolution requires combining both types with reasoning mechanisms to coordinate them. We propose a cooperative system that coordinates traffic interactions by communicating to both driver and pedestrian. The system analyzes the traffic situation, calculates cooperative behavior options, derives recommendations for each partner, and harmonizes external and internal signals. We evaluated the system in a multiplayer virtual reality experiment, pairing pedestrians and drivers in unsignalized crossing situations. Results showed significantly more interactions resolved by pedestrian crossing first, with drivers showing stronger, earlier yielding cues that let pedestrians initiate crossing faster. Additionally, the external HMI enhanced perceived driver yielding.