Pre-instruction for Pedestrians Interacting Autonomous Vehicles with an eHMI: Effects on Their Psychology and Walking Behavior (2303.08380v1)

Abstract: eHMIs refers to a novel and explicit communication method for pedestrian-AV negotiation in interactions, such as in encounter scenarios. However, pedestrians with limited experience in negotiating with AVs could lack a comprehensive and correct understanding of the information on driving intentions' meaning as conveyed by AVs through eHMI, particularly in the current contexts where AV and eHMI are not yet mainstream. Consequently, pedestrians who misunderstand the driving intention of the AVs during the encounter may feel threatened and perform unpredictable behaviors. To solve this issue, this study proposes using the pre-instruction on the rationale of eHMI to help pedestrians correctly understand driving intentions and predict AV behavior. Consequently, this can improve their subjective feelings (ie. sense of danger, trust in AV, and sense of relief) and decision-making. In addition, this study suggests that the eHMI could better guide pedestrian behavior through the pre-instruction. The results of interaction experiments in the road crossing scene show that participants found it more difficult to recognize the situation when they encountered an AV without eHMI than when they encountered a manual driving vehicle (MV); in addition, participants' subjective feelings and hesitations while decision-making worsened significantly. After the pre-instruction, the participants could understand the driving intention of an AV with eHMI and predict driving behavior more easily. Furthermore, the participants' subjective feelings and hesitation to make decisions improved, reaching the same criteria used for MV. Moreover, this study found that the information guidance of using eHMI influenced the participants' walking speed, resulting in a small variation over the time horizon via multiple trials when they fully understood the principle of eHMI through the pre-instruction.