What question did this study set out to answer?

The aim is to assess machine learning models for predicting aircraft engine failures using the C-MAPSS dataset.

February 22, 2026Open Access

Predictive maintenance in aircraft engine maintenance using the C-MAPSS dataset: performance comparison and evaluation of machine learning classification algorithms

Key Points

The aim is to assess machine learning models for predicting aircraft engine failures using the C-MAPSS dataset.
Evaluated 10 machine-learning models and 3 deep-learning models for predictive maintenance.
Used Min–Max scaling and handled class imbalance with Synthetic Minority Over-sampling Technique.
Hyperparameter tuning performed via GridSearchCV/BayesSearchCV, with performance metrics like accuracy and ROC–AUC.
LSTM achieved accuracy of 0.981 and ROC–AUC of 0.96, showing strong predictive capability.
GRU reached accuracy of 0.975 and ROC–AUC of 0.97, tracking failure dynamics effectively.
LightGBM achieved an accuracy of 0.972, demonstrating competitive performance among classical models.

Abstract

Abstract This study assesses classification-based predictive maintenance (PdM) for aircraft engines on the NASA Commercial Modular Aero-Propulsion System Simulation dataset and addresses the lack of wide-scope, unified benchmarks. PdM is cast as a short-term binary task – predicting whether an engine will fail within the next 30 cycles – and a comparison is conducted across 10 machine-learning models (Logistic Regression, Decision Tree, Random Forest, Support Vector Machine, k-Nearest Neighbor, Naïve Bayes, Extreme Gradient Boosting, LightGBM, CatBoost, and Gradient Boosting) and 3 deep-learning models (Multilayer Perceptron, Gated Recurrent Unit, and Long Short-Term Memory). A leakage-aware pipeline applies Min–Max scaling; class imbalance is handled with Synthetic Minority Over-sampling Technique where appropriate; hyperparameters are tuned via GridSearchCV/BayesSearchCV; and performance is reported with accuracy, precision, recall, F1-score, and receiver operating characteristic–area under the curve (ROC–AUC), complemented by Shapley Additive Explanations (SHAP) explainability and nonparametric significance tests. Sequence models delivered the strongest performance: LSTM achieved Accuracy = 0.981 (Macro-F1 = 0.92; ROC–AUC = 0.96), and GRU achieved ROC–AUC = 0.97 with Accuracy = 0.975 . Among classical learners, LightGBM reached Accuracy = 0.972 (Macro-F1 = 0.86; ROC–AUC = 0.93). These gains over weaker baselines were statistically significant across folds. Framing PdM as near-term failure classification yields operationally interpretable alerts. Models that explicitly capture temporal dependencies (GRU/LSTM) best track short-horizon failure dynamics, while gradient-boosted trees offer competitive, lightweight alternatives. The benchmark and analysis (including SHAP) provide a reproducible reference for model selection in aviation PdM.

Bookmark

View Full Paper

Cite This Study

Hikmetcan Özcan (Thu,) studied this question.

synapsesocial.com/papers/699a9d27482488d673cd2ebc https://doi.org/https://doi.org/10.1017/s0890060426100249

Bookmark

View Full Paper