Investigating the Need for Explicit Communication of Non-Yielding Intent through a Slow-Pulsing Light Band (SPLB) eHMI in AV-Pedestrian Interaction

Abstract

External human-machine interfaces (eHMIs) support automated vehicles (AVs) in interacting with vulnerable road users such as pedestrians. eHMI research has mostly dealt with investigating the communication an AV’s yielding intent, but there is little insight into how (or if) an eHMI should communicate an AV’s non-yielding intent. We conducted a video-based study (N = 25) with two eHMI concepts that offer pedestrians information about the vehicle’s non-yielding intent either explicitly or implicitly, and compared it with a baseline of an AV without an eHMI. Results show that while both kinds of eHMIs are effective and perform better than the baseline, there is no evidence of significant difference in road-crossing decision performance between explicit and implicit eHMIs in ambiguous situations. However, subjective feedback shows a trend of preference for eHMIs that communicate an AV’s intent explicitly at all times, although with a need for a significant distinction between the yielding and non-yielding messages.