Citation

AlMaqbali, Said and Marhoubi, Asmaa H. and Ali, Oualid (2025) A Machine Learning-Based Decision Aid for Predictive Maintenance Using Fault Diagnosis Data. In: 18th International Symposium on Embedded Multicore/Many-core Systems-on-Chip, MCSoC 2025, 15 December 2025 - 18 December 2025, Singapore.

Text IEEE Xplore Full-Text PDF_6.pdf - Published Version Restricted to Repository staff only Download (694kB)

Abstract

The proposed work explores how machine learning methods can be used as a decision support in predictive maintenance of engineering systems. Based on an example fault diagnosis dataset, we ran a simplified classification pipeline that consisted of preprocessing, feature selection, and imbalance management. Random Forest, XGBoost, and Logistic Regression were compared as three models to determine which model was the most suitable when it comes to fault-detecting tasks. The pipeline included the following features: resampling of data with SMOTE, reduction of features with ANOVA F-tests, and multiclass re-evaluation based on the macroaveraged ROC analysis. The generated experimental results confirmed that the highest accuracy of the Random Forest of 86.4% and AUC of 0.96 was achieved, then XGBoost with an accuracy of 62.0% and AUC of 0.87 followed. Logistic Regression did not perform as well, as it has an accuracy of 22.4% and AUC of 0.53, which makes it ineffective in cases of nonlinear fault detection. These results indicate that an ensemble-based model, especially the Random Forest, is suitable for decision support of predictive maintenance, which provides a high level of classification accuracy and strong diagnosis data. This study can be added to the currently existing literature on the topic of AI-based maintenance systems, as the results of the study indicate that the efficiency of industrial decision-making can be improved through the optimization of model selection

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Predictive maintenance, fault diagnosis
Subjects:	Q Science > Q Science (General) > Q300-390 Cybernetics
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	17 Mar 2026 02:19
Last Modified:	17 Mar 2026 02:19
URII:	http://shdl.mmu.edu.my/id/eprint/15461

Downloads

Edit (login required)