A prototype analysis of trust in autonomous vehicles (AVs)

Source abstract

Trust in artificial intelligence (AI), particularly autonomous vehicles (AVs), plays a critical role in shaping public policy and fostering public acceptance. As AV technology advances, measuring trust becomes increasingly important for informing government strategies. Trials currently undertaken in North America, Europe and Asia highlight the need for urban planning and transport policy departments to actively enhance public trust in AVs. This study applies prototype analysis, a technique widely used in psychology, yet to become part of the standard toolkit in transport research, to investigate the features of trust in AVs. Prototype analysis is a novel ground-up methodology to investigate the cognitive structure of an abstract concept, here trust in AVs. Rather than transferring human-human trust measures to trust in AVs, this study elicits context-specific constructs from a series of five studies undertaken in Australia: feature generation, ranking/rating, best-worst, reaction time, and vignette analysis. A key outcome is a set of policy recommendations for how government departments can implement practices that can build trust in this future technology, ensuring public confidence keeps pace with technological advancements in autonomous transport.