Study and Analysis of Human Discomfort in Autonomous Vehicles

Autonomous vehicles (AVs) are heralding a new era in transportation systems, attracting extensive attention from both academia and industries. Presently, predominant efforts are directed toward enhancing safety and efficiency in AV operations. However, the crucial aspect of human discomfort, integral to the AV user experience and capable of influencing user acceptance and eventual AV deployment, remains relatively understudied. To bridge this gap, this paper delves into the influential factors of human discomfort in AVs through human-in-the-loop studies utilizing a high-fidelity autonomous driving simulator featuring six-degrees-of-freedom motions. The investigation examined the impacts of various significant factors, including AV maneuvers, driving styles, and road types, on human discomfort. Through the application of multivariate analysis of variance and mixed logit modeling techniques, the data was quantitatively analyzed. The findings revealed the relationship between various autonomous driving factors and human discomfort, furnishing invaluable insights for the design of future autonomous vehicles.