Highly automated vehicles' communicated intents to interact with pedestrians and pedestrians' appraisals - A study across different cultures

Abstract

Highly automated vehicles (HAVs) introduce new challenges for safe pedestrian interaction, particularly in how vehicles communicate intent. This intent can be conveyed implicitly through vehicle kinematics or explicitly via external human-machine interfaces (eHMIs), such as light bands. Yet research comparing how these signals work together across cultures remains limited - knowledge that is essential for universal design and standardization. To address this gap, we replicated earlier experimental studies from Germany, Japan, and Singapore with new samples from the Netherlands and UK. Participants viewed HAVs that varied in explicit communication (intention-based eHMI, static eHMI, none) and implicit behavior (yielding, non-yielding). We measured pedestrians' willingness to cross, perceived safety, and trust in the eHMI. Across all five countries, intention-based light-bands increased willingness to cross and perceived safety, and were rated as more trustworthy and supportive. However, when explicit and implicit cues conflicted, participants tended to rely on the explicit signal, creating potentially unsafe situations. The degree of this overreliance varied slightly between national samples. Overall, the findings highlight both the benefits and risks of light-band eHMIs and reinforce the need for universal, standardized design guidelines that improve safety while reducing overreliance on explicit signals.