Light-based communication of automated vehicles with other traffic participants - a usability study in a Virtual Reality environment

Abstract

Communication with light signals is one design option that is currently discussed in the automotive domain to enable automated vehicles to interact with surrounding traffic participants. A challenge in mixed traffic is, that there are many ambiguous scenarios. These scenarios cannot only be solved with help of traffic rules. In those scenarios automated vehicles (without an active driver on board) will need to provide surrounding traffic participants communication cues to allow a common interaction planning. This paper presents the results of a Virtual Reality study investigating the interaction of an automated vehicle with a pedestrian. 27 participants took part in the study. Six different light-based designs for external human-machine-interfaces (eHMI) were tested. The eHMI comprises of a LED Band installed at the outside of the vehicle and a directed single lamp [Willrodt, 2018] installed behind the windshield in the position of the rear mirror. The eHMI was used to present information for a perception-based versus an intention-based versus a combination of an intention- and perception-based design. Three different urban scenarios covering an intersection scenario, a crossing scenario and a parking scenario were presented to find out whether participants prefer different eHMI strategies for different scenarios. Study results on the understandability and intuitiveness of the different HMI strategies in these scenarios are reported.