EEG-Based Mental Workload Estimation.

Area of Science:

• Human-Computer Interaction
• Cognitive Engineering
• Neuroscience

Background:

• Assessing operator mental workload is crucial for safety and efficiency, especially in high-demand roles like pilots and drivers.
• Existing mental workload evaluation techniques have limitations, making accurate assessment challenging.
• Persistent high mental workload can negatively impact operator performance and safety.

Purpose of the Study:

• To evaluate operator mental workload in long-duration tasks.
• To develop and validate a novel approach for continuous mental workload assessment.
• To improve task allocation and operator capability assessment.

Main Methods:

• Experiments were conducted in a simulated working environment with varied task structures (simultaneous, paused, cross-functional).
• Continuous data recording across different operational modes.
• Segmentation of recorded data for analysis of mental workload changes.
• Application of an artificial neural network (ANN) for workload classification.
• Brain connectivity analysis to validate the approach.

Main Results:

• The artificial neural network (ANN) model achieved a high classification accuracy of 96.6% for mental workload.
• The proposed approach demonstrated effectiveness in identifying and classifying mental workload levels.
• Brain connectivity analysis corroborated the efficacy of the developed method.

Conclusions:

• The study presents a highly accurate method for evaluating mental workload in long-duration tasks.
• The findings suggest that ANN-based analysis of continuous data can effectively monitor operator cognitive states.
• This approach has significant implications for optimizing operator performance, safety, and task management in demanding professions.