Cross-Platform Validation of Pedestrian Simulators

• Applicant: Professor Dr. Klaus Bengler
• Subject Area: Human Factors, Ergonomics, Human-Machine Systems ; Traffic and Transport Systems, Intelligent and Automated Traffic
• Term: from 2016 to 2021
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 317326196

• Final Report Year: 2021

Final Report Abstract

Virtual reality (VR) has become a popular tool to investigate pedestrian behavior in safe and controlled environments. In this context, Cave Automatic Virtual Environments (CAVEs) and head-mounted displays (HMDs) have emerged as the two primary technologies to display virtual traffic scenarios. Since these two types of simulators typically differ with regard to a number of technological properties, experimental outcomes may be affected by the respective hardware. In a first within-subject study, a CAVE and an HMD were compared with regard to a common gap acceptance scenario. 90 participants out of three different age groups were instructed to cross a two-lane road. Vehicles approached at a constant speed of either 40 or 60 km/h and maintained gaps between one and five seconds. More gaps were accepted in the HMD. Although in this condition, participants started crossing earlier, the acceptance of small gaps resulted in shorter safety margins and more collisions. Differences in comparison to the CAVE were especially evident in the youngest group of twelve-year-olds. In a second study, both simulators were compared to a test track. In each experimental environment, 30 young adults decided whether the gap between two cars was sufficient to safely cross a single-lane street. Again, gap sizes ranged between one and five seconds, but cars approached at a lower speed of 30 or 50 km/h. For safety reasons, no actual crossings were performed, but participants communicated their decision by stepping forward. In contrast to the first experiment, acceptance rates were similar on the test track and in the CAVE, but lower in the HMD. Furthermore, participants in both simulators were more sensitive to the approaching vehicles’ levels of speed. To assess the applicability to further technologies, the second experiment was replicated in a an augmented reality (AR) simulator. While standing on the test track, 13 participants observed virtual vehicles that were superimposed on a video stream of the environment. Matching previous observations in the HMD, acceptance rates were lower in AR than on the test track. A series of further experiments was performed in the HMD, including research on tech- niques to mitigate distance compression and the effects of scenario design. In addition to technological factors, we thereby followed a comprehensive approach to evaluate behavioral validity in pedestrian simulators. Although VR represents a valuable and flexible methodology, our results underline several challenges that must be accounted for when extending the results of simulator studies to real-world pedestrian behavior.

Publications

• (2019). “Measuring egocentric distance perception in virtual reality: Influence of methodologies, locomotion and translation gains”. In: PloS one 14.10.
  Maruhn, P., Schneider, S., and Bengler, K.
  
• (2019). “Pedestrian Behavior in Virtual Reality: Effects of Gamification and Distraction”. In: Proceedings of the Road Safety and Simulation Conference. (Iowa City, Iowa)
  Schneider, S., Ratter, M., and Bengler, K.
  
• (2020). “Analyzing Pedestrian Behavior in Augmented Reality — Proof of Concept”. In: 2020 IEEE Conference on Virtual Reality and 3D User Interfaces (VR), pp. 313–321
  Maruhn, P., Dietrich, A., et al.
  
• (2020). “Evaluating behavioral validity in traffic simulators”. In: Proceedings of TRA2020, the 8th Transport Research Arena: Rethinking transport – towards clean and inclusive mobility. (Helsinki, Finland). Ed. by F. Transport and C. A. Traficom. Traficom Research Reports, p. 130
  Schneider, S. and Bengler, K.
  
• (2020). “HMD and CAVE simulators to study pedestrian street-crossing decisions”. In: Driving Simulation & Virtual Reality Conference 2020 Europe. (Antibes, France)
  Pala, P. et al.
  
• (2020). “Virtual Scenarios for Pedestrian Research: A Matter of Complexity?” In: Lecture Notes in Computer Science 12190. HCII 2020. Ed. by J. Y. C. Chen and G. Fragomeni, pp. 171–190
  Schneider, S. and Li, G.
  
• (2020). “Virtually the same? Analysing pedestrian behaviour by means of virtual reality”. In: Transportation Research Part F: Traffic Psychology and Behaviour 68, pp. 231–256
  Schneider, S. and Bengler, K.
  
• (2021). “Analysis of Street Crossing Behaviour: Comparing a CAVE Simulator and a Head-Mounted Display among Younger and Older Adults”. In: Accident Analysis & Prevention
  Pala, P. et al.
  
• (2021). “Pedestrian Crossing Decisions in Virtual Environments: Behavioral Validity in CAVEs and Head-Mounted Displays”. In: Human Factors: The Journal of the Human Factors and Ergonomics Society
  Schneider, S., Maruhn, P., Dang, N.-T., et al.