How do the levels of automation in flight operations affect pilots' cognitive workload, reaction time, and EEG brain waves in cruising flights?

With the aid of autopilots, pilots may gradually become 'out-of-the-loop'. They may lose the capability to cope with unforeseen accidents in time. It is an opportunity to rethink how to optimise flight safety through human-automation teaming. This research evaluates the effects of levels of automation on subjective cognitive workload, reaction time, and EEG spectral features. Three levels of automation, including full automation, semi-automation, and flying manually, are evaluated. Twelve cadet pilots performed three flights, each for a level of automation since cruising, on an Airbus A320 flight simulator. EEG, reaction time, and NASA-TLX were recorded during the flight to evaluate human performance and subjective workload. Spectral analyses were performed on the EEG recordings of the frontal (Channel F4) and parietal lobe (Channel CP6). The reaction time of pilots is significantly lower in semi-automated flight than in baseline and manual flying, while the subjective cognitive workload is significantly higher in manual flying. EEG spectral features also revealed that pilots' sleepiness and drowsiness during a long time idle could be reduced significantly with semi-automated flight. Considering the trade-off between task performance and the associated workload, semi-automated flight shall keep pilots in the control loop without incurring a surge in cognitive workload. This research gives implications for future human-automation teaming in flight operations and a roadmap towards adaptive automation to keep pilots in the flight control loop.