The Effectiveness of eHMI Displays on Pedestrian-Autonomous Vehicle Interaction in Mixed-Traffic Environments

Area of Science:

• Human-Computer Interaction
• Autonomous Vehicle Safety
• Traffic Psychology

Background:

• External human-machine interfaces (eHMIs) are crucial for communication between autonomous vehicles (AVs) and road users.
• Existing research on eHMIs primarily focuses on one-to-one interactions, leaving a gap in understanding their effectiveness in complex multi-pedestrian environments.

Purpose of the Study:

• To evaluate the impact of various eHMI displays on pedestrian behavior in realistic scenarios.
• To identify optimal eHMI configurations for enhancing pedestrian safety around AVs.

Main Methods:

• A mixed-methods approach combining controlled outdoor experiments with pedestrian participants (N=31 initial, N=14 follow-up).
• An intercept survey involving 171 additional participants to gather broader perspectives.
• Exposure of participants to diverse eHMI displays during simulated AV crossing scenarios.

Main Results:

• The combination of a flashing green LED, robotic sign, and countdown timer proved to be the most effective eHMI display.
• This configuration significantly increased pedestrians' willingness to cross and reduced their response times.
• Participants demonstrated a strong preference and enhanced understanding of AV intentions with this eHMI setup.

Conclusions:

• The findings provide critical insights into designing effective eHMIs for AVs.
• This research can guide policymakers and manufacturers in developing safer urban mobility solutions.
• Optimized eHMIs are essential for fostering trust and ensuring pedestrian safety in mixed traffic environments.