EEG Dataset Collection for Mental Workload Predictions in Flight-Deck Environment

Area of Science:

• Cognitive Neuroscience
• Human-Computer Interaction
• Aerospace Psychology

Background:

• High mental workload impairs performance, particularly in high-risk professions like aviation.
• Existing deep learning models require extensive, annotated datasets for accurate cognitive state assessment.
• There's a need for specific data on mental workload and brain function in flight-deck scenarios.

Purpose of the Study:

• To present a novel dataset of electroencephalogram (EEG) recordings for mental workload recognition.
• To facilitate the development of AI models for detecting and assessing cognitive states.
• To bridge the gap in understanding brain functionality under varying mental workload conditions.

Main Methods:

• Collected EEG data from participants undergoing induced mental workload across three experiments.
• Utilized N-back test, Heat-the-Chair game, and an Airbus320 flight simulator for workload induction.
• Validated the dataset by correlating task difficulty with self-perception, performance, and EEG patterns.

Main Results:

• Demonstrated significant differences in EEG temporal patterns corresponding to varying theoretical workload levels.
• Confirmed the dataset's utility for training and evaluating artificial intelligence models.
• Established a correlation between task difficulty, self-reported workload, and objective performance metrics.

Conclusions:

• The presented EEG dataset is valuable for advancing research in mental workload assessment.
• This resource supports the development of AI-driven solutions for monitoring pilot cognitive states.
• Improved understanding and detection of mental workload can enhance aviation safety and performance.