Uncertainty feedback as component in the interaction design for highly automated vehicles

Kurzfassung

The automobile industry is working on concepts to make so called highly automated driving possible. That means that the vehicle automation takes over longitudinal and lateral control of the vehicle, and the driver must no longer permanently monitor the system. A prerequisite for this is that in situations in which the automated system might reach its limits or shows non adequate behaviour the driver can be brought back in the loop within a certain time buffer. A concept to shift the driver's attention to the driving task is to give him information about the automations uncertainty in the adequateness of its environment interpretation. In a between subjects driving simulator study with 50 participants we investigated which effects information of automation uncertainty has on the behavioural level of the driver, on drivers trust in the system and on the mental model the driver has about the system. Results indicate that uncertainty information while driving leads to more adequate mental models of the system and adjusted trust in automation, when compared to conventional alarm strategies.