predictive model for highly efficient helicopter maintenance in the royal thai air force using deep learning

The Royal Thai Air Force has helicopters in service to support tactical transport missions. Over time, helicopters deteriorate, making maintenance essential to maintain mission capabilities. Regular and timely maintenance helps to maintain operational readiness and safety, reduces the risks associated with unexpected failures, and ensures the continuous availability of critical resources. In this research, the ultimate goal is to use the results of this research as a guideline for improving the Royal Thai Air Force’s helicopter maintenance plans to be good and efficient, where efficient maintenance planning can significantly reduce costs and enhance safety. This research provides valuable insights for academia and aviation industry professionals. The researcher has proposed a model to predict helicopter maintenance in the Royal Thai Air Force to improve maintenance efficiency and increase the accuracy of spare parts calculations. Using a helicopter maintenance dataset from January 2017 to September 2020, a total of 3,819 datasets covering a variety of maintenance scenarios and operating conditions, the researchers applied deep learning (DL) techniques to make predictions. The algorithms used in this research include fully connected neural networks (FCNN), long-term short-term memory (LSTM), and convolutional neural networks (CNN). FCNN is suitable for general numerical data that are not related in sequence or space, making it effective for linear or simple numerical datasets. On the other hand, LSTM is ideal for analyzing time-sequence data because it can capture past trends to predict future outcomes. CNN excels in handling spatially correlated data, especially those related to helicopter maintenance patterns that require analyzing multiple related factors. The results show that FCNN achieves an accuracy and precision of 1.00, while both LSTM and CNN achieve an accuracy and precision of 0.94. The results of this study clearly highlight the potential of DL-based models to improve prediction accuracy. However, the study’s limitations may lie in the accuracy of deep learning model in predicting the Royal Thai Air Force’s helicopter maintenance. The future direction could be to develop more accurate predictive maintenance guidelines for the Royal Thai Air Force’s helicopters. The specific research gap this study, by improving deep learning algorithms and collecting more diverse data from the Royal Thai Air Force’s helicopters maintenance, resulting in increased accuracy. To use the results of this research as a guideline for improving The Royal Thai Air Force’s helicopter maintenance plans to be good and efficient. This will indirectly result in reducing the helicopter maintenance budget of the Royal Thai Air Force and increasing the reliability of the operations of the Royal Thai Air Force.