The Crowd Effect: How Does Pedestrian Density on Public Sidewalks Influence Discomfort Toward Autonomous Mobile Robots

This study examines how pedestrian density, lateral distance, and external Human-Machine Interface (eHMI) type influence discomfort and avoidance behavior with autonomous mobile robots (AMRs). A within-subjects virtual reality experiment with 40 participants manipulated three pedestrian density levels, four lateral distances, and three eHMI types (none, cyan, or red projection). Discomfort and discomfort distance were analyzed using generalized linear mixed-effects models. Higher pedestrian density increased perceived discomfort while reducing discomfort distance, whereas narrower lateral distances increased both outcomes. eHMI type influenced perceived discomfort but not discomfort distance, with the red projection producing greater discomfort. Interaction effects revealed a critical spatial threshold around 120 cm, where the influence of pedestrian density on perceived discomfort shifted from amplification to attenuation. This demonstrates a dissociation between affective and behavioral responses, suggesting constraints may prevent pedestrians from acting on discomfort. The findings highlight the need for density-aware navigation and proactive lateral spacing in shared environments.