MECHANISM OF DYNAMIC ADAPTATION OF SIMULATOR PARAMETERS FOR IMPLEMENTING INDIVIDUAL-AND-TEAM STRATEGIES IN AIR TRAFFIC CONTROLLER TRAINING

Abstract

The paper presents mechanism for dynamic adaptation of simulator content aimed at implementing individual-and-team strategies in air traffic controller (ATCO) training within a competency-based learning paradigm. The relevance of the study is determined by the transition of ATCO training to ICAO- and EASA-based competency models, as well as by the need to improve simulator efficiency under increasing traffic complexity, staffing constraints, and stringent flight safety requirements. The purpose of the paper is to design mechanisms for dynamic control of simulator parameters that adapt the scenario to the individual-and-team state of trainees. The research combines systems analysis, competency-based training and assessment principles, Bayesian Knowledge Tracing for estimating the probability of mastering specific skills, and interpretable machinelearning models for explaining the team result. A cyclic six-module architecture of an adaptive simulator is proposed. It includes behaviour monitoring, diagnostics of risk and good-practice indicators, prediction of skill acquisition, dynamic scenario adaptation, explainable assessment, and post-session reporting. The study formalizes the session state vector, workload index, skill-deficit index, and the rule of wave-like complexity variation based on the logic “increase – peak – recovery”. It is shown that adaptation should be not only quantitative but also targeted: depending on the dominant deficit, the system selects a focused event from a library of scenario interventions without driving the team beyond the productive training zone. Gradient boosting models combined with SHAP analysis are proposed to explain team assessment and to identify individual, negative, positive, and compensatory contributions of team members. The practical value of the results lies in creating a foundation for a new generation of simulator systems capable of maintaining trainees in a productive workload zone in real time, generating a posttraining report for instructor, and increasing the precision of corrective pedagogical influence