Correlation between Expected Workload and EEG Indices of Cognitive Workload and Task Engagement

As a Human Computer Interaction (HCI) designer, it is important to understand the cognitive workload level (mental demand) of an operator while performing a task. A particular concern about operator workload stems from the increased potential for errors when cognitive workload gets too high. The Office of Naval Research Disruptive Technologies Opportunity Fund, is sponsoring research investigating using neurophysiological data to develop a more accurate measurement of cognitive workload than metrics currently available. As a part of this research effort, an HCI usability evaluation was conducted which involved having 9 Navy operators interact with a prototyped Tactical Tomahawk Weapons Control System (TTWCS) while wearing several neurophysiological sensors. More specifically, fleet operators were asked to use the prototype to respond to system status and prompts. Their objective was to launch each of 32 missiles on time. Data from the evaluation were analyzed to identify the correlation between EEG-generated measures of cognitive workload, measures of predicted workload identified by Multimodal Information Design Support (MIDS) and an expert’s observed rating of the operator’s cognitive workload. Correlations of EEG based measures of cognitive workload averaged across all participants and the MIDS ‘standard’ expected workload values (not customized for individuals) were statistically significant. However, correlations of EEG based measures of cognitive workload for each participant and Expert ratings of each participant’s cognitive workload were not statistically significant. The latter finding may be due to insufficient data. While these findings give a first indicator of promise for an EEG-based measure of cognitive workload in an HCI environment, further research addressing limitations of the study will be needed. In particular, precision timesynchronization and logging of all participant activities are essential to meaningful interpretations of EEG data.