Pedestrians' responses to scalable automated vehicles with different external human-machine interfaces: Evidence from a video-based eye-tracking experiment
Wei Lyu,Wen-gang Zhang,Xueshuang Wang,Yi Ding,Xinyue Yang
DOI: https://doi.org/10.1016/j.trf.2024.04.005
IF: 4.349
2024-04-15
Transportation Research Part F Traffic Psychology and Behaviour
Abstract:To enhance the efficiency and safety of interactions with pedestrians, numerous external Human-Machine Interfaces (eHMIs) concepts for automated vehicles (AVs) have been proposed and evaluated, predominately based on singular pedestrian-AV interaction scenarios. This leaves a gap in comprehending the efficiency and robustness of eHMIs during interactions with scalable AVs. To bridge the gap, this study pioneers an exploration of pedestrians' road-crossing decisions, perceived clarity, and gaze behaviour during synchronous interactions with multiple AVs equipped with different eHMIs. An eye-tracking experiment was conducted, involving 48 video stimuli depicting two AVs (with yielding patterns and eHMIs manipulated) approaching a predetermined road-crossing location from the same side of a two-way, four-lane, un-signalized road. Each yielding AV will present one of four eHMI manipulations (i.e., light band, smiling expression, pedestrian symbol, and no eHMI as baseline), resulting in 16 unique eHMI combinations and video stimuli. Eighty-seven participants were recruited and tasked to respond to approaching AVs, during which their road-crossing decision time, perceived clarity, and gaze metrics were recorded. Findings revealed that different eHMIs across AVs were associated with prolonged road-crossing decision times, diminished perceived clarity, and intensified visual attention as evidenced by increased fixation counts and durations. Regarding AVs equipped with the same eHMIs, the pedestrian symbol-based eHMIs resulted in heightened clarity perception, reduced crossing decision times and visual engagement, in contrast to eHMIs presented in abstract light bands or smiling expressions. Furthermore, for eHMI-different AVs, those featuring only one eHMI within AVL-AVR pairs were perceived as less clear, eliciting heightened visual cognitive load and decision times. Integration of a pedestrian symbol-based eHMI in AVL-AVR configurations garnered superior clarity ratings and facilitated time savings. These findings of this pioneering study provide insights for eHMI design for future AVs in complex traffic scenarios with scalable road users.
transportation,psychology, applied