Designing for Calibrated Trust Towards Operators of Highly Autonomous Vehicle Systems

Abstract: Appropriate levels of trust in highly autonomous vehicles contributes to a safer human-machine relationship. Through increased correctness in interactions trust can be calibrated. Trust is formed via interactions that needs to be designed from a Human-Machine relationship with the purpose of maintaining it at an acceptable level. When failing calibrating trust to an appropriate level accidents and risks may arise due to over-reliance on, or the opposite, distrust on the system's capabilities. This study aims to investigate what factors of trust that can influence design decisions and the design process in order to let user's trust levels match the system's capabilities (i.e. calibrating trust) from a human-machine interface (HMI) perspective. The method of semi-structured expert interviews have given a basis to stand on in an idea generation workshop. The main results from this study are various ways to achieve transparency which in turn can be used to design for appropriate trust with the help of the generated concepts.