Trust-oriented HMI Design for Conditional Autonomous Driving Takeover Systems

Conditional Autonomous Driving (CAD) will become widespread in the next few years, but the driving features of human-vehicle co-driving could mean significant safety and trust issues. A good cooperation between human and vehicles requires the driver to trust the vehicle appropriately, otherwise it will lead to misuse or disuse of the vehicle by the driver. In this study, based on the theoretical model of trust in automation through literature review, we conducted semi-structured interviews with eight drivers (male=4, female=4) to collect drivers' information needs in the full phase of CAD and preference of existing HMI design. The data was organized and coded based on the interview results. We sorted out the priority of the first 6 HMI messages in pre-takeover, in-takeover and after-takeover stages respectively. The findings were applied to the visual (dashboard and center control screen) and audio HMI in typical takeover scenarios to increase driver's trust in the autonomous driving takeover system and improve driver acceptance and security perception of CAD vehicles.