Estimating Individual Growth Processes During Pilot Training Using Software Reliability Growth Models
Kento Yamada, Harumi Ikeshita, Yuta Kyoya, Makoto Ueno · 2023
This paper aims to classify and quantify individual growth processes during pilot training to understand the variations in their growth processes. Various software reliability growth models were fitted to growth processes, and the best-fit model, which showed the minimum Akaike information criterion, was clarified for each applicant. As a result, it was shown that the individual growth processes were classified into concave and S-shaped processes. The concave and S-shaped processes were attributed to the cases where an applicant completed the training without and with wandering, respectively. Since the best-fit models included models with imperfect debugging and testing efforts these factors were discussed from the viewpoint of flight training. It was also indicated that the estimated growth models after the training can be utilized to monitor the applicant's competencies depending on operation-service flights done up to recurrent training.