Real-Time Mental Workload Estimation Using EEG

Tracking mental workload in real-time during worker’s performance is a challenge, as it requires the worker to report during the task execution. Moreover, it is thus based on subjective experience. Neuroergonomics tackles this issue, by employing neurophysiological metrics to obtain objective, real-time information. We measured mental workload (MWL) derived from electroencephalography (EEG) signals, for subjects engaged in a simulated computer-based airplane-landing task. To test this metric, we calculated the degree of correlation between measured MWL and observable variables associated to task complexity. In the two settings of the experiment, we used a 24-channel full-cap EEG system and the novel mobile EEG headphone device. The latter allows seamless integration of the EEG acquisition system in a possible real-world setup scenario. Obtained results reveal significant correlation between the EEG derived MWL metric and the two objective task complexity metrics: the number of airplanes on the screen subjects had to control, as well as the number of actions performed by the subject during the task in both setups. Therefore, this work represents a proof of concept for using the proposed systems for reliable real-time mental workload tracking.