, Associate Professor, UMD-Computer Science
Following a snowplow (or any vehicle) during snowy conditions can be extremely dangerous. This danger lies in the inability of the human visual system to accurately perceive the speed and motion of the snowplow ahead, often resulting in rear-end
collisions. For this project, the researchers used their understanding of how the human visual system processes optical motion under the conditions created by blowing snow to create a simulation framework for testing emergency lighting configurations that could reduce rear-end collisions with snowplows. Reaction times for detecting the motion of the snowplow were measured empirically for a variety of color set-ups on a simulated
snowplow that slowed down while driving on a virtual road with curves and hills. The researchers implemented a blowing snow model that will eventually be integrated into a real-time driving simulation environment. Concurrently, a simulated driving environment was developed to serve as the basis for testing the effects of color and lighting alternatives on snowplows. In initial pilot experiments, the simulated driving environment was effective for testing subject reaction times for following a snowplow through high-luminance contrast (normal daylight driving) and low-luminance contrast (daylight fog) conditions. The results of this work will move the researchers closer to determining optimal color and lighting configurations on actual snowplows.