Evaluation of Visual Performance from Pedestrian Crosswalk Lighting

Crosswalk lighting is important because a substantial number of nighttime traffic-related fatalities involve pedestrians. Current recommendations for crosswalk lighting are based upon horizontal or vertical illuminance levels, but illuminance levels are not necessarily predictive of the visibility of pedestrians for drivers at night. Four pedestrian crosswalk lighting configurations were evaluated in an outdoor experiment in conjunction with low-beam vehicle headlamps: no fixed lighting, pole-mounted lighting directly over the crosswalk, pole-mounted lighting offset from the crosswalk, and a bollard lighting system. Adult- and child-sized, black silhouettes were used as targets. Subjects identified the orientation of a pedestrian target (facing left or facing right) on each trial as quickly as possible. The bollard-based system resulted in the shortest identification times. Importantly, the identification times were not well correlated with illuminance levels, but were strongly correlated with predictions of the speed and accuracy of detection based upon a visual performance model that considers observer age, target size, target contrast and background luminance.