Autonomous Driving Experience (ADX) design is an emerging field of research and design that centres on enhancing the passengers’ user experience in autonomous vehicles. The key is the provision of transparent communication regarding the perceptions and intentions of these autonomous vehicles. This thesis is grounded in the principles of experience design. It applies Human Computer Interaction research
methods to investigate autonomous driving experiences. Specifically, it explores how using uni- and multimodal feedback modalities can positively influence passengers’ user experience, their perceived feeling of safety and control, and their understanding of the autonomous driving system. These factors are crucial for the successful adoption of autonomous vehicles.
Autonomous driving relies heavily on artificial intelligence (AI) for reasoning and decision-making. However, passengers may feel uneasy and out of control without adequate explanations. This research addresses this issue by delving into Explainable AI and exploring how, when, and in what contexts specific aspects should be explained or visualised to enhance the passengers’ experience.
The research unfolds in several stages. Initially, an extensive and systematic literature review reveals the need for a more human-centred approach to experience design in autonomous driving. Subsequently, four experiments were conducted. These experiments investigated the impact of different feedback modalities for transparent communication on the passengers’ user experience, perceived feeling of safety and control, and system understanding. The first three experiments demonstrate their positive influence. The fourth experiment, however, uncovers the limitations of transparency communication in non-critical failure situations, necessitating alternative approaches. In the next phase, all recommendations and contributions generated throughout this comprehensive research endeavour were distilled and encapsulated within a card-based toolkit - the ADX cards. These cards serve as a valuable resource for the field, providing a tangible and accessible embodiment of the research’s insights. To ensure the validity and practical utility of the ADX cards, the toolkit underwent an iterative design process, expert evaluations, and extensive workshop testing. Furthermore, the ADX cards were successfully applied in four distinct real-world case studies, further validating their effectiveness and relevance, as well as this thesis’ research.
The main contributions of this research are twofold. Firstly, the experiments shed light on the critical role of transparency communication in enhancing the autonomous driving experience and establish a foundation to do so. Secondly, the knowledge acquired throughout this thesis has been distilled into a card-based toolkit, the ADX cards, making the research findings readily accessible and applicable to those interested in advancing the field of autonomous driving experience.