Towards pre-emptive physiological error classification within airborne environments

Tools

McGuire, Niall and Greene, Jacob and Moshfeghi, Yashar; Oliver, Nuria and Shamma, David A. and Candello, Heloisa and Cesar, Pablo and Lopes, Pedro and Artizzu, Valentino and Draxler, Fiona and Lopez, Gustavo and Reinschluessel, Anke V. and Tong, Xin and Toups Dugas, Phoebe O., eds. (2026) Towards pre-emptive physiological error classification within airborne environments. In: CHI EA '26: Proceedings of the Extended Abstracts of the 2026 CHI Conference on Human Factors in Computing Systems. Association for Computing Machinery (ACM), ESP. ISBN 9798400722813 (https://doi.org/10.1145/3772363.3798722)

[thumbnail of McGuire-etal-2026-Towards-pre-emptive-physiological-error-classification-within-airborne-environments]
Preview
 Text. Filename: McGuire-etal-2026-Towards-pre-emptive-physiological-error-classification-within-airborne-environments.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo
Download (1MB)| Preview

Abstract

Human error contributes to 70-80% of aviation accidents despite technological advances, yet current safety systems remain reactive. We present the first empirical validation that neurophysiological signals can predict pilot errors up to 10 seconds in advance during operational flight. Through trials with nine commercial pilots, we collected synchronised EEG and eye-tracking data across laboratory, straight-and-level flight, and 2G manoeuvre conditions whilst participants performed multitasking scenarios. EEG-based classifiers achieved peak F1 scores of 88.7% in controlled environments and 82.0% during flight operations, with eye-tracking providing complementary predictive patterns. Results demonstrate that physiological error precursors remain detectable despite motion artefacts, electromagnetic interference, and gravitational stress. These findings establish the feasibility of pre-emptive error monitoring in operational aviation environments, providing a foundation for cognitive-aware cockpit systems that could enable proactive safety interventions before errors occur.

ORCID iDs

McGuire, Niall ORCID logoORCID: https://orcid.org/0009-0005-9738-047X, Greene, Jacob and Moshfeghi, Yashar ORCID logoORCID: https://orcid.org/0000-0003-4186-1088; Oliver, Nuria, Shamma, David A., Candello, Heloisa, Cesar, Pablo, Lopes, Pedro, Artizzu, Valentino, Draxler, Fiona, Lopez, Gustavo, Reinschluessel, Anke V., Tong, Xin and Toups Dugas, Phoebe O.
CORE (COnnecting REpositories)