You Walkin' to Me? How Footstep Sound Primes Anticipation in Virtual Pedestrian Collision Avoidance

Studying pedestrian navigation in crowded environments is key to many applications, like crowd simulation for urban planning, but is challenging under real-world conditions due to the lack of experimental control. This motivates the use of virtual reality to manipulate variables and recreate complex social interactions. While research has mostly emphasized visual feedback, and eventually haptic rendering of collision, the integration of auditory cues in VR platforms, such as footstep sounds, remains overlooked. The present study investigates the role of adding footstep sounds in a VR collision-avoidance task. Twenty participants completed a within-subject experiment where a virtual pedestrian, initially visually occluded, crossed their path either with or without audible footsteps. Locomotor adjustments and gaze behavior were analyzed before and during the visual interaction. Results showed that footstep sounds induced earlier gaze orientation and more proactive trajectory adjustments from the participants, without affecting their walking speed. In contrast, the absence of sound led to more collisions and crossing order inversions, indicating reduced efficiency. These findings demonstrate that simple auditory cues can guide attention, facilitate anticipation, and improve interactions with virtual agents. Integrating sound into VR platforms is therefore essential to study pedestrian behavior and social navigation.