Safety at first sight? Influence of external human-machine-interfaces for highly automated vehicles

Kurzfassung

The development of highly automated vehicles (AV) and their integration into road traffic is seen by many car manufacturers as a revolution in mobility. However, this new type of mobility raises questions about the AV’s interaction with its environment. Regardless of a human driver, AVs must be able to manoeuvre safely on the road and communicate with road users in their environment. Latest research discusses external human-machine interfaces (eHMI) as a possible solution. In addition to the positive effects of eHMI, however, possible negative effects on road safety must also be considered. Method: Using the example of a road crossing pedestrian in an urban context, this experimental study examines possible positive and negative effects of eHMIs on highly automated vehicles. In addition to the timing and trustworthiness of the crossing decision, the focus is also on the frequency of safety related gaze checks of the participant to secure the road crossing. Further, a rating regarding the usability of the eHMI was assessed. In a VR study, 52 test persons had to cross a road populated with mixed traffic, which included non-automated vehicles as well as automated vehicles (with vs. without eHMI). Results: The study results show that test persons felt significantly safer in their decision to cross the road when the AV was equipped with an eHMI. In addition, they crossed the road significantly earlier when the AV was equipped with an eHMI. However, the frequency of gaze checks did not differ significantly between the different vehicles. Conclusions: Overall, results indicate that AVs with an eHMI stand out from other vehicles in mixed traffic by presenting additional information. Negative influences of these additional information via the eHMI on the frequency of gaze control could not be demonstrated. Participants reported to be well-informed by the eHMI and that the displayed information was easy to perceive. Furthermore, the information on the eHMI helped participants to anticipate the future behaviour of the AV and increased their feeling of safety. Implications for a holistic evaluation of eHMI and its safe introduction into road traffic will be discussed.