Context-Dependent Assessment of Driving Dynamics and User Perception in Automated Driving

Automated driving systems are increasingly operating in complex urban environments, where driving behavior is shaped not only by technical requirements but also by how users perceive the vehicle’s motion. While many studies evaluate automated driving performance in isolated maneuvers or specific traffic situations, little empirical evidence exists on whether similar driving dynamics are perceived consistently across different urban contexts. This work investigates whether comparable motion patterns, such as acceleration profiles or steering behavior, are evaluated similarly or differently when occurring in roundabouts and intersections using data from real-world driving studies. The empirical basis consists of two user studies. In the first study, 24 participants experienced automated driving in roundabout scenarios in public traffic. In the second study, 20 participants experienced automated driving through intersections on a test track under defined boundary conditions. In both studies, high-resolution vehicle data capturing longitudinal and lateral dynamics were recorded and linked to subjective ratings of comfort, safety, and perceived driving quality. To enable a direct comparison across contexts, the vehicle data were processed and segmented, and a set of motion-related characteristic parameters was derived. These parameters capture relevant aspects of longitudinal and lateral vehicle dynamics and allow comparable motion patterns to be analyzed across both scenarios. Statistical analyses were applied to examine the relationships between these metrics and subjective evaluations, and to identify systematic, context-dependent differences. The results show that similar dynamic features can lead to differing perceptions of driving quality depending on the urban context, particularly between roundabouts and intersections. These findings provide a data-driven foundation for context-aware evaluation of automated driving behavior and support targeted improvements in motion planning and control strategies to enhance perceived driving quality.