Smart evaluation of instrument scan pattern using state transition model during flight simulator training

Trainee pilots are expected to be thoroughly trained on both technical and nontechnical skills. Technical skills are relatively easy to evaluate. However, non-technical skills are hard to monitor and assess. This study investigates the feasibility of providing a smart evaluation technique that can generate feedback on trainee pilots’ instrument scan behavior. The authors conducted gaze monitoring experiments with a number of trainee pilots while operating a flight simulator in order to isolate patterns of behavior associated with Situation Awareness (SA). Recorded data on eye movements are then processed and transitions are extracted using a state transition model. The sequence of transitions are then processed to retrieve repeated instrument scan pattern. The results are verified using chart visualizations. © Springer International Publishing AG 2018.