HMI and VR for trustable and comfortable driving on autonomous vehicles: Using simulation to compare different driving styles

Autonomous driving technology has received extensive attention in the continuous popularization of automation, and is now getting rather mature. However, before fully enjoying the advantages of autonomous driving, users must trust autonomous vehicles (AVs), just as a prerequisite for using automation. The motion behavior of AVs, i.e., their driving style, is one of the keys to achieving the users' trust. Hence, the goal of this thesis is to develop a framework that can be used to study how AVs' driving style affects user trust, with the ultimate objective of using potentially achievable findings to program automation. To reach this goal, state-of-the-art in the field is first reviewed with the aim of summarizing the influencing factors of trust in the process of human-machine interaction, and filtering out the elements that can be transmitted through the driving style to enhance trust. Afterward, an existing VR-based synthetic environment is used to simulate various driving styles, from a "defensive" to an "aggressive" one. The used environment simulates the complex and realistic behavior of traffic and pedestrians. The environment supports motion simulation. The exploited vehicle behaves according to SAE Level 5. The differences in driving styles are mainly reflected in acceleration, overtaking, and braking behavior in selected, pre-configured driving situations. The thesis's outcomes will be allowed to perform the experiments requested for the next developments.