Application of predictive maintenance models for aircraft engine reliability and safety enhancement: a case of civil aviation in Uzbekistan

This study proposes a short-horizon predictive maintenance (PdM) model that predicts turbofan engine failures within five flight cycles using routine sensor data. A Random Forest classifier trained on 2,800 synthetic cycles achieved strong performance (ROC-AUC = 0,88), confirming the value of thermal and vibration indicators. The results show that reliable failure prediction is possible even with limited data and support the integration of AI-based diagnostics into MSG-3 and CAMO practices in Uzbekistan’s civil aviation.