A neuroergonomics approach to investigate the mental workload of drivers in real driving settings

Abstract

The safety and performance of automobile drivers depend on many factors. The mental status of the drivers is the foremost factor in ensuring driving safety, in addition to physical elements. Studies with drivers are generally conducted in driving simulators or with scenarios close to actual driving. This study investigated the mental workload of drivers by analyzing electroencephalography data recorded in totally spontaneous real driving tasks, to determine the effect of different road conditions and driving experience. Two mental workload indexes (Frontal Theta/ Parietal Alpha and Frontal Midline Theta) and a mental fatigue index (Alpha + Theta/Beta) were calculated using the band powers. Drivers were found to experience a higher mental workload in road sections with heavy traffic and variable road parameters by analyzing EEG data in real traffic. The correlation coefficients between the fatigue index and the two workload indexes were found to be 0.577 and 0.678, respectively. The workload decreased with increasing driving experience. Therefore, having experienced drivers perform commercial driving tasks can ensure safer driving. By employing novel methods to handle real-world EEG data, autonomous driving systems can also benefit.