Sleep debt and disruptions to daily biological rhythms, linked to various factors (long shifts, irregular working hours, night flights, etc.), require pilots to constantly push their biological limits. Furthermore, pilots’ mental workload varies significantly during a flight. Whilst it is high during critical phases (take-off and landing), it drops significantly during cruise phases. When the mental workload becomes too high or, conversely, too low, performance deteriorates and piloting errors may occur. In this context, the implementation of methods to detect drowsiness and mental workload levels in near real time presents a major challenge for monitoring and controlling piloting activities.
