Detection of Pilot's Mental Workload Using a Wireless EEG Headset in Airfield Traffic Pattern Tasks

Area of Science:

• Aviation Psychology
• Neuroscience
• Human Factors Engineering

Background:

• Elevated mental workload (MWL) in pilots can impair performance and compromise flight safety.
• Real-time detection of pilot MWL is crucial for proactive safety interventions.

Purpose of the Study:

• To develop and evaluate a functional system for identifying and detecting pilots' real-time MWL.
• To assess the effectiveness of a wireless headset-based system in a flight simulator environment.

Main Methods:

• Utilized a realistic flight simulator with designed airfield traffic pattern tasks.
• Assessed perceived MWL using NASA Task Load Index (NASA-TLX) scores.
• Extracted physiological features via Fast Fourier Transformation and performed feature selection using Kruskal-Wallis (K-W) test and Sequential Forward Floating Selection (SFFS).
• Analyzed electroencephalography (EEG) features and power spectral density (PSD) changes across different MWL levels.
• Employed 10-fold cross-validation on six classifiers, with a multi-class K-Nearest Neighbor (KNN) classifier yielding optimal results.

Main Results:

• Optimal input features were identified as all PSD features.
• The multi-class KNN classifier achieved an optimal accuracy of 87.57% in classifying different MWL levels.
• Demonstrated the reliability and feasibility of the wireless headset-based system.

Conclusions:

• The developed wireless headset system effectively detects pilots' real-time MWL.
• The system shows potential for application in various real-world driving and aviation scenarios.
• This research contributes to the development of future systems for monitoring cognitive states in operational environments.