Can Pedestrians Judge The Time-to-Arrival of Approaching Vehicles?

Abstract

Purpose: : The aim of this study was to determine how well normally-sighted pedestrians can judge the time-to-arrival of approaching vehicles using either vision and hearing or hearing only.

Methods: : Twenty-nine normally sighted subjects stood with their eyes closed 0.5m before a curb of an unsignalized street. Subjects wore earbud-style headphones attached to a laptop computer which played white noise. At random times, an audible tone was given, at which time the white noise paused and subjects observed an approaching vehicle for two seconds under one of two sensory conditions: (i) habitual vision and habitual hearing or (ii) masked vision and habitual hearing. During the two second observational time, the habitual vision and habitual hearing subjects opened their eyes such that they could "see" and "hear" the approaching vehicle. The masked vision and habitual hearing subjects kept their eyes closed and so could only "hear" approaching vehicles. At the end of this period a second tone was given, at which time the white noise resumed and subjects closed their eyes and recorded the vehicle’s perceived time-to-arrival by pressing a hand held trigger; each trigger press corresponded to the estimation of one second of the vehicle’s travel time. The error, computed as the difference between the actual and perceived time-to-arrival, was calculated for each subject under both sensory conditions and an analysis of co-variance was used to determine if the error changed significantly as a function of sensory condition and time-to-arrival.

Results: : Our preliminary analysis of the data showed that the error changed as a function of time-to-arrival (t₅₁₃₃=44.76, p<0.001). The way in which the error changed as a function of time-to-arrival however was different for the two sensory conditions (time-to-arrival x sensory condition interaction, t₅₁₃₂=-13.55, p<0.001). We found that with increasing time-to-arrival, the amount of under-estimation of the actual time-to-arrival was greater under the hearing only condition than it was under the vision and hearing condition.

Conclusions: : Our data suggests that the further away the vehicle was from the pedestrian, the greater the inaccuracy at judging the time-to-arrival. Pedestrians tended to under-estimate the time-to-arrival of approaching vehicles by a greater amount when using only auditory information compared to using both vision and hearing.

Keywords: low vision • vision and action