Influences of Personal Driving Styles and Experienced System Characteristics on Driving Style Preferences in Automated Driving

Abstract

As automated driving technology continues to advance, the question of how users prefer to be driven in their new, more passive role is becoming increasingly relevant. In this paper, a real-world study on a German motorway with 42 participants was conducted to analyze driving style preferences for conditional automated driving, taking the participants’ personal driving style into account. In the first part, participants’ personal driving style (PDS) was recorded during a manual drive in the first half on a given route. For the second half, participants were asked to demonstrate their desired driving style (DDS) for conditional automated driving. In the second part, participants were driven on the same route in a defensive automated vehicle (AV) while rating driving comfort and safety. Subsequently, the relationship between driving style differences and ratings was analyzed. Furthermore, a comparison between PDS and DDS was performed. The results show that very defensive to moderate drivers perceived the AV’s driving style, being similar to their own, as equally safe but significantly more comfortable than moderate to very aggressive drivers. No influence of driving style differences was found on the increase in trust. However, a significant increase in trust after experiencing an automated vehicle has been observed. Furthermore, the rated system characteristics of anthropomorphism, safety, and overall driving strategy had a significant influence on driving style preferences for AVs. This study makes an important contribution to answering the question of how users want to be driven in conditional automated driving.