Carvatar: Increasing Trust in Highly-Automated Driving Through Social Cues

While driving in highly automated new problems occur which are not present in manual driving. Highly automated cars have different internal states which take into account, what was detected, were the car is and to which maneuvers this could lead. One problem is that the driver may not be able to see in which state the car is currently because the functionality is too complex. If drivers are not able to compare the car's actions to the actions they would have performed themselves they might have trouble trusting the system. It is not apparent what the car recognizes and which future actions are planned. Our avatar uses social cues and anthropomorphism to translate the car's state into human behavior and expressions which can be interpreted intuitively by the driver. The driver is therefore more aware of the situation and might gain more trust in the system.