2025/07/28 by Marcus, Viktoria, Pitts, Griffin, Motamedi, Sanaz · 1 citation
#Computers and Society (cs.CY) #Emerging Technologies (cs.ET) #FOS: Computer and information sciences #Human-Computer Interaction (cs.HC)
paper · doi:10.48550/arxiv.2507.21303
Each year, over half of global traffic fatalities involve vulnerable road users (e.g. pedestrians), often due to human error. Level-5 automated driving systems (ADSs) could reduce driver errors contributing to pedestrian accidents, though effectiveness depends on clarity and understandability for other road users. External human-machine interfaces (eHMIs) have been proposed to facilitate pedestrian-ADS communication, though consensus on optimal eHMI features remains unclear. In an online survey, 153 participants responded to road-crossing scenarios involving level-5 ADSs, with and without eHMIs. With eHMIs, pedestrians crossed earlier and more confidently, and reported significantly increased perceptions of safety, trust, and understanding when interacting with level-5 ADSs. Visual eHMI features (including a text display and external speedometer) were ranked more necessary than auditory ones, though auditory cues received positive feedback. This study demonstrates that eHMIs can significantly improve pedestrians' understanding of level-5 ADS intent and enhance perceived safety and trust, facilitating more intuitive pedestrian-ADS interactions.