Dynamic access to the actual driver state

Abstract

As an inadequate level of workload is an important contributing factor to accidents, information management and driver assistance systems (ADAS) are developed with the general aim to optimize driver workload. This is done either by controlling the amount of information presented to the driver or by supporting parts of the driving task by an ADAS. As the workload of the driver changes with different driving tasks in a highly dynamic traffic these systems have to react in a similar dynamic manner. This requires a fast and dynamically changing estimation of the driver’s workload. In recent real driving studies we could show that driver workload changes with different driving manoeuvres and additionally depends on environmental factors. There were also indications that workload varied even within one manoeuvre. To examine workload influenced by different driving manoeuvres, different phases within the manoeuvres and different environmental factors a simulator study was conducted where different traffic densities were varied and driver workload was measured continuously. This is used to analyse the level and time-course of workload depending on several influencing factors. Implications of these results for the adaptation of ADAS to the workload of the driver are discussed.