Visual Information for Car Following by Drivers: Role of Scene Information

Abstract

An important task during driving is the maintenance of headway during car following. The visual information available to a driver for successful car following was examined. A model of car following that used the visual angle and change in visual angle of a lead vehicle was developed. The study examined whether information from the surrounding scene (e.g., the roadway and buildings) influenced car-following performance. Licensed drivers were presented with a car-following task in a driving simulator. The simulator display consisted of a roadway scene with a lead vehicle that varied speed according to a sine wave. To evaluate the role of scene information, car-following performance was examined when the surrounding scene was present or absent. Two frequencies and three amplitudes of speed variation were also examined. The results indicated that control gain was greater when the scene was absent, with near unity gain when the scene was present. These findings indicated more accurate control during car following when the surrounding scene information was present. These results suggest that drivers also use other sources of information (e.g., absolute distance information from height in the visual field relative to the horizon and edge rate information specifying observer speed) to maintain headway. The implications of this research to nighttime accident rates and intelligent highway systems are discussed.