Presentation
Consistency Between Implicit Cues and eHMIs in Autonomous Vehicles that Accelerate and Decelerate: Effects on Pedestrian Crossing Decisions
SessionThursday Poster Session
DescriptionExplicit communication cues that can be understood without inferences (e.g., a driver’s gestures) are absent from driverless automated vehicles (AVs), which only provide implicit cues (e.g., acceleration). Although implicit cues may be sufficient for crossing decisions, prior research indicated that pedestrians made faster decisions when explicit cues were available from external human-machine interfaces (eHMIs). However, it is unknown how explicit cues that contradict implicit cues (i.e., malfunctions) affect pedestrians’ crossing decisions and their trust, particularly when the AV displays large (i.e., presumably perceivable) acceleration patterns (acceleration, deceleration, none). In the current study, an AV approached pedestrians while decelerating, accelerating, or maintaining its speed, with an eHMI that presented an explicit cue that was consistent or inconsistent with its motion behavior. Participants decided whether to cross the road. Results showed that participants based their crossing decisions on both implicit and explicit cues, even when the eHMI provided an inconsistent explicit cue (e.g., “WALK” when the AV accelerated). Similarly, trust of the AV was greater when implicit and explicit cues were consistent than when they were inconsistent. Findings suggest that pedestrians considered both implicit and explicit cues, emphasizing the need for consistent cues in AVs to avoid confusion in their crossing decisions.
Event Type
Poster
TimeThursday, October 22nd5:30pm - 6:30pm PDT
Location
Similar Presentations