Contribution
Assessing Pedestrian Behavior in Traffic: A Pilot Study on Real-World Time to Contact Estimations
* Presenting author
Abstract:
Virtual reality has proven effective for investigating pedestrian perception in street-crossing scenarios, offering controlled and reproducible experimental conditions. To quantify the pedestrian behavior in traffic scenarios, the Time to Contact point (TTC) was previously investigated in other studies, by simulating approaching vehicle scenarios with setups such as Wave Field Synthesis (WFS) and Higher Order Ambisonics (HOA). In order to validate these findings, it is essential to conduct a real-world TTC experiment. To this end, a pilot study, where an electric vehicle approached participants at varying constant speeds, was realized. The study was conducted in an auditory-only manner, with participants wearing virtual reality glasses that visually masked the participants and displayed a graphical user interface created in MATLAB. This interface enabled them to record their TTC estimates by pressing a button at the moment they anticipated the vehicle would arrive. To allow the perception of real-world auditory cues, participants were equipped with open headphones, which permitted the transmission of the approaching vehicle's sounds while simultaneously introducing noise to mask the vehicle sound at designated instances. The results demonstrated reasonable TTC estimates, offering insight into both the effective aspects of the experimental setup and potential areas for improvement in future designs.