Havacılık bakım, onarım ve yenileme sektöründe yapay zeka uygulamalarının çkkv yöntemleri ile seçilmesi

Airline operators are looking for ways to improve flight performance and flight safety, and to minimize maintenance-repair-overhaul (MRO) costs and the number of unplanned breakdowns over time. It is exceedingly difficult to achieve optimal results in such a large system with hundreds of variable factors. However, technological developments facilitate the exchange of data between interrelated operational activities and make it meaningful by processing the big data that emerges as a result of the operations performed. In this sense, artificial intelligence (AI) and machine learning concepts have started to be a big supporter of the MRO companies, such as in predictive and preventive maintenance issues, data processing, reporting activities and forecasting incidents etc. In this thesis study, it will be carried out to determine the most appropriate area in which AI technology can be used in aviation MRO activities and to detect the most suitable AI tool for this determined area. In the first stage of the study, the potential processes for which AI technology can be used in MRO operations was found out. The WEDBA method, which is one of the multi-criteria decision-making methods, was used to determine which of these potential processes is more suitable for this technology. The next step is to investigate which AI technology tools can be used in the specified process. The most appropriate AI tool for this specified process is determine by the WEDBA method again. For both applications, criterion weighting is performed by AHP method. As a result of the study carried out, eight potential fields where the technology can be applied in, and 11 different criteria were determined to decide which of these fields would be more applicable. In this study, it has been determined that the "predictive & preventive maintenance" is the most suitable area according to the result of WEDBA implementation. After this study, a comprehensive market research was conducted for predictive and preventive maintenance applications, and 11 prominent software programs that offer solutions for this field were identified. 13 criteria have been determined to evaluate these software programs while performing the WEDBA method. After all, "Alternative-5" out of 11 alternative AI tools was found to be the most suitable alternative for the implementation of predictive and preventive maintenance in aviation MRO sector. For the further studies, sensitivity analysis can be made through different MCDM methods, and the results can be compared, a feasibility analysis for the determined implementation can be made to investigate whether such an investment is financially feasible or not, and possible project risks can be revealed with a comprehensive risk analysis.