Evaluation for Trainee Pilot Workload Management Competency During Approach Phase Based on Flight Training Data

With the implementation of furthering the flight training reform, competency-based training has drawn much attention for the initial flight training in aviation schools. The competencies assessment is the key to conduct the competency-based training program which is originated from the evidence-based training. The workload management (WLM) competency of pilots, which belongs to the non-technical competency, plays an important role to the flight safety. At present, most methods to evaluate the workload management competency are usually carried out through the questionnaire afterward or subjective score from the instructor's observation. However, a variety of information from the flight record data which objectively represent the flight operation behaviors has not been fully exploited. In this paper, the assessment approach for the WLM was developed based on the flight training data. Consider the correlation and redundancy of the high-dimension temporal data, the principal component analysis (PCA) was adopted to deal with the dimension reduction and feature extraction firstly. Then, the processed flight record data and the instructor score were utilized to conduct the supervised learning with the deep neural network (DNN). The case study for the approach phase WLM evaluation of trainee pilots was given to demonstrate the effectiveness of the proposed method, and the results show the differences between the prediction score through the WLM evaluation and the actual score from instructors are fairly close, which proved the feasibility and efficiency of the proposed method for the WLM evaluation.