Analysis of drivers control methods of task demands based on behavioral data in actual road environments

This paper describes an investigation of how drivers control task demands while approaching intersections in actual road environments. Field experiments using an AIST instrumented vehicle were conducted on rural roads (Tsukuba area) and urban roads (Tokyo area) to collect data on vehicle status and driver behavior. We collected driver behavior data while driving in their usual manner and while driving with low task demands. A Bayesian network model was applied to the data while approaching and stopping at intersections with red traffic lights. We investigated the behavioral indices that indicated a significant difference between the usual drives and the drives with low task demands. The model estimated results suggest that the drivers control the task demands using vehicle velocity, headway distance, accelerator pedal application, or brake pedal application. The behavioral indices indicating the differences are the same on both the rural and urban roads.