Virtual Reality Study on Automated vehicles – Pedestrians Interaction

With the advent of highly automated vehicles, traditional driver-pedestrian interactions based on gestures or eye contact will disappear, and new solutions will be needed to convey vehicles’ intentions. To ensure the safest interaction between pedestrians and automated vehicles, some critical questions need to be addressed first, like for example, what information is necessary and relevant for the pedestrians? Or, how should the information be communicated to ensure drivers understand it correctly?

Recently, some of these questions have been addressed in a study as part of the BRAVE project. This study aimed to investigate how the vehicle deceleration, a strong indicator of drivers’ intention to stop or not, should be integrated with the explicit information provided by an external human-machine interface (eHMI). By using a virtual reality walking simulator (see figure hereafter), 29 participants were presented with different automated vehicles in a crossing scenario. The vehicle decelerations and eHMI message presentations occurred at far or close distance to the pedestrians, and simultaneously (i.e., deceleration and eHMI message at the same time) or independently (i.e., early deceleration followed by a later eHMI message, or vice versa). Based on this, pedestrians had to infer the vehicle’s intention and decide whether to cross or not.

The results showed that the vehicle’s intention is significantly more evident when the eHMI is early presented and accompanied by the vehicle deceleration (i.e., eHMI and deceleration simultaneously). This indicates that the vehicle dynamic remains as a strong cue to infer the automated vehicle intentions and therefore, need to be integrated into future eHMI systems.

Experimental setting in the walking simulator at VTI, Linköping

Leave a Reply

Your email address will not be published. Required fields are marked *

Why ask?