Evaluation of the driver’s psychophysiological state based on EEG and fNIRS measurements within a driving simulator

Authors

K. Steinmetzger, Uni Heidelberg

Summary

As a result of the execution of primary driving functions by the vehicle, the driver is increasingly relieved in automated mobility. This is expected to be accompanied by a change in the focus of attention. This has a negative effect on the driver's performance in critical takeover situations. Market-ready attention assistants analyse driving behaviour to assess the current driver state. In automated means of transport, such an evaluation does not make sense. Instead, there is considerable potential in the development of contactless sensor technology that uses vital parameters and behavioural characteristics to estimate the driver's performance. To validate such systems, however, a scientifically sound reference (ground truth) is necessary. For this purpose, questionnaires or subsequent evaluations by experts are usually used in subject studies. These methods have considerable disadvantages due to subjectivity, discontinuity, and poor reproducibility.

In contrast, within the framework of this project, we are striving for an objective evaluation of the driver's state on the basis of measurements of brain activity. For this purpose, electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) are applied simultaneously while subjects are presented with different distractor stimuli in a driving simulator and the automation level is systematically varied. The EEG and fNIRS data are analysed under different modelling assumptions to identify neural markers associated with attention and performance while driving. In addition to analyses of evoked brain activity, the focus will be on oscillatory analyses and novel measures of signal complexity.