TITLE:
Cognitive Workload Assessment of Aircraft Pilots
AUTHORS:
Maxime Antoine, Hamdi Ben Abdessalem, Claude Frasson
KEYWORDS:
Cognitive Workload, Aviation, Heart Rate, Pupil Dilation, Machine Learning, Deep Learning, EEG
JOURNAL NAME:
Journal of Behavioral and Brain Science,
Vol.12 No.10,
October
10,
2022
ABSTRACT: In this research, we study the cognitive workload of aircraft pilots during a simulated takeoff procedure. We propose a proof-of-concept setup environment to gather heart rate, pupil dilation, and brain cognitive workload data during an A320 takeoff within a simulator. Experiments were performed during which we collected 136 takeoffs across 13 pilots for more than 9 hours of time-series data. Moreover, this paper investigates the correlations between heart rate, pupil dilation, and cognitive workload during such exercise and found that a spike in cognitive load during a critical moment, such as an engine failure, augments a pilot’s heart rate and pupil dilation. Results show that a critical moment within a takeoff procedure increases a pilot’s cognitive load. Next, we used a stacked-LSTM model to predict cognitive workload 5 seconds into the future. The model was able to produce accurate predictions.