Remote Control of Autonomous Vehicles in VR Exploring the Impact of Speed on Interaction Methods in Remote Driving

Introduction As autonomous vehicles advance, human–machine interaction (HMI) in remote driving scenarios requiring manual intervention becomes increasingly critical. This thesis explores how driving speed impacts user performance, comfort, and cognitive load within a gesture-based VR remote driving environment. Research Question The primary research question is: “How does driving speed influence cognitive workload, usability, and driving performance in a gesture-based VR remote driving system?” Method Twelve licensed drivers participated in a within-subjects experiment using a Meta Quest Pro headset to control a virtual vehicle via hand gestures at three speed levels (30, 50, 70 km/h). Objective metrics (Road Edge Collisions, Obstacle Collisions) and subjective questionnaires (NASA Raw-TLX, two SUS items (ease of use, confidence), and custom experiential items) were collected. Data analysis employed repeated-measures ANOVA and Wilcoxon signed-rank tests with Holm correction. Results Results show that higher speeds increase cognitive workload, collision rates, and motion sickness symptoms, while moderate speed (50 km/h) yields optimal usability, comfort, and perceived safety. Subjective and objective metrics consistently indicate participants’ preference for moderate speeds in remote VR driving scenarios. Discussion These findings emphasize the importance of adaptive speed control in VR gesture-based remote driving interfaces to balance performance and user comfort. Limitations include the small sample size and simulated environment constraints. Future research should explore multimodal feedback integration and real-world validations to further improve system safety and user experience.