Pilot's Attention Distribution Modeling Using Hidden Markov Models

The attention distribution over the instrument panel in flight is a significant indicator of pilots cognitive state. A major difference between a rookie and experienced pilot can be observed through the means of perceiving the environment and processing the information gathered from the flight instruments. Then such attention distribution models is a part of pilot behavior modeling and even pilot identification or classification. This work is focused on fusion of the eye-tracking data with the flight variables acquired during simulated flight experiments. The flight data itself were collected in a series of experimental simulated flight tasks performed by real pilots with a different level of experience.