ABSTRACT
Investigations of accidents and incidents in highly automated commercial aircraft have identified pilot error as a major cause. In most cases they involved situation awareness (SA) errors due to failures in information perception during the pilot’s monitoring process. To describe the monitoring process in more detail a method is presented which analyses and formalises empirical data of the pilot’s scanning behaviour using a machine learning algorithm. A decision tree is developed which displays the monitoring process in a specific situational context and allows setting up detailed hypothesis about possible strategies in his visual attention control.
