Pilot stress study uses multimodal biometrics to track in-flight fatigue

A new study published on arXiv (ID: 2608.09947) by Atandrila Chowdhury, Sudip Vhaduri, Julius Keller, Debra Henneberry, and Mark Wilson examines how stress and fatigue build in student pilots during real flight training. The researchers collected continuous physiological data—heart rate (HR), electrodermal activity (EDA), skin temperature, and acceleration—throughout flight sessions, along with Perceived Stress Scale (PSS-10) self-reports before and after each flight. By analyzing physiological signals across different flight stages, they aimed to identify recurring patterns of arousal and workload that indicate cumulative strain.

The findings reveal a clear trend: both EDA and skin temperature steadily increased across consecutive flights, reflecting a buildup of workload-related weariness over time. Post-flight self-reported exhaustion scores also rose, confirming that pilots feel increasingly tired as training sessions accumulate. Heart rate, however, showed more event-specific behavior—brief spikes during high-demand phases such as landings or complex maneuvers, rather than a steady upward drift. This suggests HR captures acute stress events, while EDA and skin temperature better track longer-term fatigue. The study demonstrates the value of fusing passive physiological sensors with subjective reports to create a holistic picture of pilot well-being. It also highlights the potential of data-driven approaches for real-time monitoring, which could inform flight training schedules, detect early signs of burnout, and improve aviation safety by flagging fatigued pilots before critical mistakes occur.