Using Immersive Virtual Reality to Study Road-Crossing Sustainability in Fleeting Moments of Space and Time

Abstract

Despite a history of year-by-year reduction in road-crossing harm and fatality in the United States, the trend reversed course in 2009 and road-crossing has grown more hazardous since. Within this tendency, there has been a marked uptick in risk to urban crossers who are neither children nor elderly. The age group in between these extremes represents a bulk of urban crossers, for whom theoretical explanations for crossing behavior that are focused on youth and senior crossing factors often do not apply. New insight is likely required to explain why the rate of crossing harm is growing for the 20–44 age group, but declining among the young and elderly. However, it is difficult to experiment with crossing scenarios in a real-world context, where significant dangers are present and for which the uniqueness of crossers and crossing sites is abundant. In this paper, we introduce an end-to-end system for examining crossing behavior using a unique combination of real human crossing behavior, made safe through the combination of agent-based models, motion capture, virtual geographic environments, and immersive technologies from virtual reality. We demonstrate that this combination of methods can be deployed to examine very high resolution and very high specificities of crossing scenarios and behaviors, with reach to individual crossers and their judgment over tiny windows of space and time. We demonstrate that the system can reproduce known effects from the theoretical literature and from existing case studies, while also generating huge swaths of empirical and diagnostically useful data on crossing actions, interactions, and reactions relative to fleeting events and phenomena of urban geography, traffic dynamics, and ambient pedestrian crowds. To prove the concept, we deploy the system to investigate crossing judgment behavior among the 20–44 age group.