Assessing pedestrian safety across modalities via a simulated vehicle time-to-arrival task.

Pedestrians must use a variety of visual and auditory cues when determining safe crossing opportunities. Although vision has received a bulk of the attention in research on pedestrian safety, the examination of both vision and audition are important to consider. Environmental, intrapersonal, and cognitive qualities of a pedestrian context may limit the use of one or both perceptual modalities. Across two experiments, we examined the impact of perceptual constraints on pedestrian safety by measuring the accuracy of vehicle time-to-arrival estimates in a virtual environment when vehicles were only visible, only audible, or both visible and audible. In both experiments, participants estimated the time-to-arrival of vehicles moving at one of two speeds (8-kph, 40-kph). In the second experiment, we introduced ambient traffic noises to examine the impact of environmentally relevant traffic noises on pedestrian perception. Results suggest seeing a vehicle is more advantageous than hearing a vehicle when interacting with traffic, especially in the presence of ambient sound. Both experiments resulted in more accurate time-to-arrival estimates in the visual and mixed conditions than in the auditory-only condition. Implications for pedestrian safety and future research are discussed.