A Trustworthy Framework for Condition Monitoring and Remaining Useful Life Prediction of Rotating Machinery

Authors: Umme Habiba Aesha; Reyan Hridoy Bhuiyan; Yeasir Arafat Ayan
A Trustworthy Framework for Condition Monitoring and Remaining Useful Life Prediction of Rotating Machinery
DOI
10.25125/ijoer-jul-2026-2
DIN
IJOER-JUL-2026-2
Abstract

The increasing adoption of smart manufacturing technologies has intensified the need for reliable predictive maintenance solutions to reduce unexpected equipment failures and production downtime. Rotating machinery, including bearings, electric motors, gearboxes, pumps, and turbines, represents a critical class of industrial assets whose degradation directly affects manufacturing productivity, operational safety, and maintenance costs. This study presents a trustworthy AI-driven framework for condition monitoring and Remaining Useful Life (RUL) prediction of rotating machinery in smart manufacturing environments. The proposed framework integrates multi-sensor condition monitoring using vibration, temperature, acoustic emission, pressure, and motor current measurements with machine learning and deep learning models for intelligent fault diagnosis and prognostics. To improve transparency and industrial trust, the framework incorporates Explainable Artificial Intelligence (XAI) techniques, including SHAP and LIME, to identify the operational factors influencing prediction outcomes. In addition, Digital Twin technology and Physics-Informed Artificial Intelligence are integrated to enhance prediction reliability and maintain consistency with engineering knowledge. The framework is evaluated using benchmark datasets and standard classification and regression metrics, including Accuracy, Precision, Recall, F1-Score, MAE, RMSE, and RUL prediction error. The results demonstrate that the integration of multi-sensor monitoring, explainable AI, and Digital Twin-assisted analysis improves diagnostic reliability, prediction consistency, and maintenance decision support. The proposed approach offers practical benefits for Industry 4.0 applications by reducing unplanned downtime, optimizing maintenance scheduling, improving equipment availability, and enhancing the trustworthiness of AI-based predictive maintenance systems for critical rotating machinery.

Keywords
Trustworthy Artificial Intelligence Predictive Maintenance Condition Monitoring Remaining Useful Life Prediction Rotating Machinery Smart Manufacturing Industry 4.0 Explainable Artificial Intelligence Digital Twin Physics-Informed AI Machine Learning Deep Learning.
Download Options
Download Full PDF
Share Article
Facebook Twitter LinkedIn
Introduction

The integration of digital technologies into production systems drives the transformation of the manufacturing industry. Intelligent manufacturing connects components such as factory machines and sensors to collect and analyze operational data. This transformation, tied to Industry 4.0, shifts production maintenance from passive response to data-driven decision-making. According to references [1]–[6], this transformation can improve operational efficiency, equipment utilization, and production reliability. Among the assets operating within modern manufacturing facilities, rotating machinery plays a particularly important role. Bearings, electric motors, gearboxes, pumps, and turbines provide the mechanical power required for production processes across sectors such as automotive manufacturing, chemical processing, power generation, mining, and oil and gas. Because these components often operate continuously under demanding mechanical and environmental conditions, gradual degradation is unavoidable. If developing faults remain undetected, they may lead to unexpected equipment failure, production downtime, costly repairs, and increased safety risks [7]–[12]. To reduce these risks, industries have increasingly adopted condition monitoring as part of predictive maintenance programs. Rather than relying solely on fixed maintenance schedules, condition monitoring assesses equipment health using operational measurements collected during machine operation. Signals such as vibration, temperature, acoustic emission, pressure, and motor current contain valuable information about the mechanical condition of rotating equipment. They can reveal early indications of wear, imbalance, misalignment, lubrication problems, or bearing damage before catastrophic failure occurs [13]–[20]. Recent advances in artificial intelligence have significantly expanded the capabilities of condition monitoring systems. Machine learning and deep learning models can identify complex relationships in sensor data that are difficult to detect with conventional analytical methods. These models have demonstrated considerable success in machinery fault diagnosis and Remaining Useful Life (RUL) prediction, allowing maintenance activities to be planned before equipment reaches a critical failure state. Accurate RUL prediction contributes to improved maintenance scheduling, reduced operational costs, and increased equipment availability, making it an important component of intelligent maintenance strategies [21]–[30]. Although predictive performance has improved substantially, industrial deployment of AI-based maintenance systems still faces practical challenges. Many high-performing models provide limited explanation for their predictions, making it difficult for maintenance engineers to verify whether recommended actions are technically reasonable. This lack of transparency can reduce confidence in automated maintenance decisions, particularly for safety-critical equipment. Consequently, recent research has focused on developing trustworthy AI by incorporating explainability, engineering knowledge, and Digital Twin technologies to improve the reliability and interpretability of intelligent maintenance systems [2], [30]. Motivated by these developments, this study investigates a trustworthy AI-driven framework for condition monitoring and Remaining Useful Life prediction of rotating machinery in smart manufacturing environments. The proposed framework combines multi-sensor data analysis with explainable artificial intelligence and Digital Twin concepts to support reliable maintenance decision-making, improve equipment availability, and reduce unplanned production interruptions.

Engineering Journal IJOER Call for Papers

Conclusion

This study proposes a trustworthy artificial intelligence framework for rotating machinery in intelligent manufacturing scenarios. The framework integrates six technologies, enables condition monitoring and remaining useful life prediction, and supports reliable, transparent predictive maintenance. By combining data-driven intelligence with engineering-informed analysis, the framework addresses the growing demand for maintenance systems that not only deliver high predictive performance but also provide interpretable, trustworthy decision support for industrial applications. The study demonstrated that integrating vibration, temperature, acoustic emission, pressure, and motor current measurements enables comprehensive assessment of equipment health and improves the detection of machinery degradation. The incorporation of machine learning and deep learning techniques enables accurate fault diagnosis and Remaining Useful Life estimation. At the same time, Explainable Artificial Intelligence enhances model transparency by identifying the operational factors that influence prediction outcomes. Furthermore, integrating Digital Twin technology with Physics-Informed AI enhances prediction reliability by combining real-time operational data, engineering knowledge, and virtual asset representations. The proposed framework offers several practical benefits for modern manufacturing industries. By enabling early fault detection and accurate Remaining Useful Life prediction, it supports condition-based maintenance strategies that reduce unplanned downtime, optimize maintenance scheduling, lower maintenance costs, and improve equipment availability. The addition of explainable and engineering-informed decision support further increases user confidence, making the framework more suitable for deployment in safety-critical industrial environments where maintenance decisions must be technically justified. Despite these contributions, several opportunities remain for future research. Further experimental validation using large-scale industrial datasets and real-world manufacturing environments will strengthen the practical applicability of the proposed framework. In addition, future studies may investigate federated learning, edge artificial intelligence, uncertainty-aware prediction models, and more advanced Digital Twin implementations to improve scalability, computational efficiency, and prediction robustness. Overall, this research advances trustworthy predictive maintenance by presenting a comprehensive framework that integrates intelligent analytics with transparent, engineering-consistent decision-making. The proposed approach supports the objectives of Industry 4.0 by improving equipment reliability, enhancing manufacturing productivity, reducing operational risks, and advancing intelligent maintenance systems for next-generation smart manufacturing.

References
  1. Antoni, J. (2006). The spectral kurtosis: A useful tool for characterising non-stationary signals. Mechanical Systems and Signal Processing, *20*(2), 282-307.
  2. Armstrong, N. (1987). Desire and domestic fiction: A political history of the novel. Oxford University Press.
  3. Beauvoir, S. de. (1989). The second sex (H. M. Parshley, Trans.). Vintage Books. (Original work published 1949)
  4. Breiman, L. (2001). Random forests. Machine Learning, *45*(1), 5-32.
  5. Brontë, C. (1847). Jane Eyre. Smith, Elder & Co.
  6. Caesarendra, W., & Tjahjowidodo, B. (2015). A review of feature extraction methods in condition monitoring and machine fault diagnosis. Measurement, *69*, 148-156.
  7. Cerrada, M., Zurita, G., Cabrera, D., Sánchez, R., Li, C., & de Oliveira, R. V. (2020). Fault diagnosis in rotating machinery using signal processing and machine learning techniques. IEEE Access, *8*, 212728-212748.
  8. Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785-794.
  9. Cho, K., et al. (2014). Learning phrase representations using RNN encoder–decoder for statistical machine translation. Proceedings of EMNLP, 1724-1734.
  10. Cortes, C., & Vapnik, V. (1995). Support-vector networks. Machine Learning, *20*(3), 273-297.
  11. Cover, T., & Hart, P. (1967). Nearest neighbor pattern classification. IEEE Transactions on Information Theory, *13*(1), 21-27.
  12. Davidoff, L., & Hall, C. (1987). Family fortunes: Men and women of the English middle class, 1780–1850. Hutchinson.
  13. Deutsch, J., & He, D. (2018). Using deep learning-based approach to predict remaining useful life of rotating components. IEEE Transactions on Systems, Man, and Cybernetics: Systems, *48*(1), 11-20.
  14. dos Santos, F. B., de Souza, R. B. M., & Torres, L. H. S. (2019). Machine learning techniques applied to predictive maintenance: A systematic literature review. Computers & Industrial Engineering, *137*, Article 106024.
  15. Eagleton, T. (1975). Myths of power: A Marxist study of the Brontës. Macmillan Press.
  16. Eagleton, T. (1976). Marxism and literary criticism. University of California Press.
  17. FEMTO-ST Institute. (2012). PRONOSTIA bearing accelerated life test data set.
  18. Gilbert, S. M., & Gubar, S. (1979). The madwoman in the attic: The woman writer and the nineteenth-century literary imagination. Yale University Press.
  19. Heng, A., Zhang, S., Tan, A. C. C., & Mathew, J. (2009). Rotating machinery prognostics: State of the art, challenges and opportunities. Mechanical Systems and Signal Processing, *23*(3), 724-739.
  20. Hinton, G. E., Vinyals, O., & Dean, J. (2015). Distilling the knowledge in a neural network. arXiv, arXiv:1503.02531.
  21. Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, *9*(8), 1735-1780.
  22. Jardine, A. K., Lin, D., & Banjevic, D. (2006). A review on machinery diagnostics and prognostics implementing condition-based maintenance. Mechanical Systems and Signal Processing, *20*(7), 1483-1510.
  23. Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017). ImageNet classification with deep convolutional neural networks. Communications of the ACM, *60*(6), 84-90.
  24. Kusiak, A. (2018). Smart manufacturing. International Journal of Production Research, *56*(1-2), 508-517.
  25. LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, *521*(7553), 436-444.
  26. Lee, J., Bagheri, B., & Kao, H.-A. (2015). A cyber-physical systems architecture for Industry 4.0-based manufacturing systems. Manufacturing Letters, *3*, 18-23.
  27. Lee, J., Wu, F., Zhao, W., Ghaffari, M., Liao, L., & Siegel, D. (2016). Prognostics and health management design for smart manufacturing systems. Manufacturing Letters, *8*, 31-34.
  28. Lei, C., Li, N., & Lin, J. (2018). Machinery health prognostics: A systematic review from data acquisition to remaining useful life prediction. Mechanical Systems and Signal Processing, *104*, 799-834.
  29. Lei, Y., Li, N., Guo, L., Li, N., Yan, T., & Lin, J. (2017). Machinery health prognostics: State of the art and future trends. Mechanical Systems and Signal Processing, *94*, 173-203.
  30. Lei, Y., Li, N., & Lin, J. (2016). A new method based on stochastic process models for machine remaining useful life prediction. IEEE Transactions on Instrumentation and Measurement, *65*(12), 2671-2684.
  31. Li, Y., Ding, X., & Liu, Z. (2020). Remaining useful life prediction using deep long short-term memory networks. Mechanical Systems and Signal Processing, *138*, Article 106570.
  32. Marx, K. (1990). Capital: A critique of political economy (Vol. 1). Penguin Classics. (Original work published 1867)
  33. Marx, K., & Engels, F. (2002). The Communist manifesto. Penguin Classics. (Original work published 1848)
  34. Mobley, R. K. (2002). An introduction to predictive maintenance (3rd ed.). Elsevier.
  35. Moglen, H. (1976). Charlotte Brontë: The self conceived. W. W. Norton & Company.
  36. Nandi, S., Toliyat, H. A., & Li, X. (2005). Condition monitoring and fault diagnosis of electrical motors—A review. IEEE Transactions on Energy Conversion, *20*(4), 719-729.
  37. Paderborn University. (2020). Paderborn bearing data center.
  38. Poovey, M. (1988). Uneven developments: The ideological work of gender in mid-Victorian England. University of Chicago Press.
  39. Randall, R. B., & Antoni, J. (2011). Rolling element bearing diagnostics—A tutorial. Mechanical Systems and Signal Processing, *25*(2), 485-520.
  40. Rhys, J. (1966). Wide Sargasso Sea. André Deutsch.
  41. Rich, A. (1979). Jane Eyre: The temptations of a motherless woman. In On lies, secrets, and silence: Selected prose 1966–1978 (pp. 89-106). W. W. Norton & Company.
  42. Rigby, E. (1848). Vanity Fair—and Jane Eyre. Quarterly Review, *84*(167), 153-185.
  43. Saxena, A., & Goebel, K. (2008). Turbofan engine degradation simulation data set. NASA Ames Prognostics Data Repository.
  44. Saxena, A., Goebel, K., Simon, D., & Eklund, N. (2008). Damage propagation modeling for aircraft engine run-to-failure simulation. International Conference on Prognostics and Health Management.
  45. Showalter, E. (1977). A literature of their own: British women novelists from Brontë to Lessing. Princeton University Press.
  46. Spivak, G. C. (1985). Three women's texts and a critique of imperialism. Critical Inquiry, *12*(1), 243-261.
  47. Tao, F., Qi, Q., Wang, L., & Nee, A. Y. C. (2019). Digital twins and cyber–physical systems toward smart manufacturing and Industry 4.0: Correlation and comparison. Engineering, *5*(4), 653-661.
  48. Vaswani, A., et al. (2017). Attention is all you need. Advances in Neural Information Processing Systems, *30*, 5998-6008.
  49. Wang, H., Liang, P., & Zhang, X. (2020). Hybrid CNN–LSTM network for remaining useful life prediction of rotating machinery. IEEE Access, *8*, 14324-14336.
  50. Wang, S., Xiang, J., Zhong, Y., & Tang, H. (2021). A data-driven intelligent fault diagnosis method for rotating machinery. IEEE Transactions on Industrial Electronics, *68*(10), 10430-10441.
  51. Wollstonecraft, M. (1988). A vindication of the rights of woman. Norton. (Original work published 1792)
  52. Xu, L. D., Xu, E. L., & Li, L. (2018). Industry 4.0: State of the art and future trends. International Journal of Production Research, *56*(8), 2941-2962.
  53. Yin, S., & Kaynak, O. (2015). Big data for modern industry: Challenges and trends. Proceedings of the IEEE, *103*(2), 143-146.
  54. Zhang, K., Tang, B., Deng, X., & Huang, W. (2021). Remaining useful life prediction based on deep learning: A survey. Reliability Engineering & System Safety, *211*, Article 107587.
  55. Zhang, P., Qin, Y., Wang, D., & Zhang, H. (2021). Intelligent condition monitoring of rotating machinery: Recent advances and future trends. IEEE Access, *9*, 155214-155236.
  56. Zhang, X., Ge, Z., & Song, Z. (2022). Data-driven monitoring technologies for smart manufacturing. IEEE Transactions on Industrial Informatics, *18*(3), 1847-1859.
  57. Zhao, Z., Liang, B., Wang, X., & Lu, W. (2019). Deep learning and its applications to machine health monitoring. Mechanical Systems and Signal Processing, *115*, 213-237.
Article Preview
Quick Actions
Download PDF View References
Article Info
Pages
10-31