Evaluating Virtual Locomotion Techniques for Pedestrian Simulation

Kurzfassung

The validation of automated vehicle systems requires realistic pedestrian behavior in virtual simulations. However, existing behavior models of vulnerable road users (VRUs) often fail to capture the complexity of human decision making. Human-in-the-Loop simulator studies offer an alternative, which can integrate real user behavior. But, pedestrian locomotion methods that promote realistic pedestrian behavior and decision making remain a challenge. In this late breaking report, we introduce a study to compare three locomotion techniques (by controller, walk-in-place, and omnidirectional treadmill) to evaluate their impact on usability, immersion, and behavioral realism. These locomotion methods will be compared across different obstacle courses and within an urban environment. Key research questions will address the influence of locomotion techniques on naturalistic behavior, decision making, presence, and simulation sickness. Our goal is to establish best practices for integrating unlimited pedestrian movement into VR simulations, improving scenario validity for traffic research and automated vehicle testing. With this report, we seek to engage in discussions on study design, movement patterns, and scenario selection to refine our approach and advance VRU behavior modeling in simulation-based validation.