A Composite Cognitive Workload Assessment System in Pilots Under Various Task Demands Using Ensemble Learning

The preservation of attentional resources under mental stress holds particular importance for the execution of effective performance. Specifically, the failure to conserve attentional resources could result in an overload of attentional capacity, the failure to execute critical brain processes, and suboptimal decision-making for effective motor performance. Therefore, assessment of attentional resources is particularly important for individuals such as pilots who must retain adequate attentional reserve to respond to unexpected events when executing their primary task. This study aims to devise an expert model to assess an operator’s dynamic cognitive workload in a flight simulator under various levels of challenge. The results indicate that the operator’s cognitive workload can be effectively predicted with combined classifiers of neurophysiological biomarkers, subjective assessments of perceived cognitive workload, and task performance. This work provides conceptual feasibility to develop a real-time cognitive state monitoring tool that facilitates adaptive human-computer interaction in operational environments.