Delphi and Bayesian Networks in the Analysis of Fatigue in Air Traffic Events in Practical ATC Instruction

Risk management is an ongoing process that involves the planning of an organization’s resources, whether human or material, aiming to reduce and/or eliminate the occurrence of certain risks, in addition to mitigating the effects of those that may happen. The objective of this work was to propose the use of the Delphi method and the Bayesian Networks in the evaluation of the influence of fatigue on occurrences of air traffic involving the process of practical instruction. To achieve this objective, a preliminary report corresponding to the exercise from February 2019 to February 2020 was considered. The report pointed out that of the 15 existing sectors that cover the airspace region within the scope of the São Paulo Operations Center/In Brazil, sectors T7, T8, and T5 (grouped with T3 and/or T4) had more occurrences. Three (03) controllers who act as instructors, assigned probability values considering (a) the influence of fatigue during practical instruction (b) the sectors involved, (c) the complexity of the tasks to be performed, and (d) the existence or not of adverse weather conditions. The results points out that the complexity of the task is the most prominent factor for fatigue and the consequent occurrence of incidents, followed by the existence of meteorological conditions and the operation in sector T5 (grouped to T3 and/or T4). The developed research reached, in general, its objectives, being able to contribute as a tool for risk management and decision making during the practical instruction process.