Home / Current Issue / Paper 1723999
Artificial Intelligence for Predictive Maintenance of Mission-Critical Engineered Systems: A Systematic Review of Methods, Frameworks, and Future Directions
Subject area: Science,Engineering and Technology · Area of research: Industry 4.0
DOI: 10.64388/IREV10I4-1723999
Abstract
Asset management in safety-critical industries is moving away from calendar-driven preventive schedules and toward autonomous, data-driven decision-making enabled by artificial intelligence (AI) and machine learning (ML). This work reports a PRISMA-compliant systematic review of the predictive maintenance (PdM) literature published between January 2018 and June 2026, covering domains such as aerospace, energy, and advanced manufacturing where component degradation must be tracked with high reliability. A structured search across four bibliographic databases (IEEE Xplore, Scopus, Web of Science, and the ACM Digital Library) produced an initial pool of 2,847 candidate records; after multi-stage screening, 127 peer-reviewed studies were retained for detailed synthesis. The retained literature is organized into six methodological threads: purely data-driven learning, physics-informed machine learning, digital-twin architectures, causal and explainable AI, uncertainty quantification, and human-AI collaboration. Although deep architectures achieve strong benchmark accuracy, the synthesis surfaces recurring shortcomings around out-of-distribution robustness, epistemic uncertainty handling, and model opacity. Eight recurring research gaps are catalogued, and a forward-looking agenda is proposed, culminating in a conceptual unifying framework, referred to here as NEXUS-AI, that combines physical priors, neuro-symbolic reasoning, and human oversight to support verifiable reliability in high-consequence settings.
Keywords
predictive maintenance; prognostics and health management; remaining useful life; physics-informed machine learning; digital twin; explainable AI; causal inference; uncertainty quantification; systematic review
References
[1] Mobley RK. An Introduction to Predictive Maintenance. 2nd ed. Amsterdam, The Netherlands: Elsevier; 2022.
[2] Jardine AKS, Lin D, Banjevic D. A review on machinery diagnostics and prognostics implementing condition-based maintenance. Mech Syst Signal Process. 2006;20(7):1483–1510.
[3] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med. 2009;6(7):e1000097.
[4] Lei Y, Yang B, Jiang X, Jia F, Li N, Nandi AK. Applications of machine learning to machine fault diagnosis: A review and roadmap. Mech Syst Signal Process. 2020;138:106587.
[5] Zhao R, Yan R, Chen Z, Mao K, Wang P, Gao RX. Deep learning and its applications to machine health monitoring. Mech Syst Signal Process. 2019;115:213–237.
[6] Saxena A, Goebel K, Simon D, Eklund N. Damage propagation modeling for aircraft engine run-to-failure simulation. In: Proc. Int. Conf. Prognostics Health Manage. (PHM); 2008. p. 1–9.
[7] Si X, Wang W, Hu C, Zhou D. Remaining useful life estimation—A review on the statistical data driven approaches. Eur J Oper Res. 2011;213(1):1–14.
[8] Goebel K, Saha B, Saxena A, Celaya JR, Christophersen J. Prognostics in battery health management. IEEE Instrum Meas Mag. 2008;11(4):33–40.
[9] Sateesh Babu G, Zhao P, Li X. Deep convolutional neural network based regression approach for estimation of remaining useful life. In: Proc. DASFAA; 2016. p. 214–228.
[10] Zheng S, Ristovski K, Farahat A, Gupta C. Long short-term memory network for remaining useful life estimation. In: Proc. IEEE ICPHM; 2017. p. 88–95.
[11] Zhang J, Wang P, Yan R, Gao RX. Long short-term memory for machine remaining life prediction. J Manuf Syst. 2018;48:78–86.
[12] Li X, Ding Q, Sun J. Remaining useful life estimation in prognostics using deep convolution neural networks. Rel Eng Syst Saf. 2018;172:1–11.
[13] Wen L, Gao L, Li X. A new deep transfer learning based on sparse auto-encoder for fault diagnosis. IEEE Trans Syst, Man, Cybern, Syst. 2019;49(1):136–144.
[14] Ince T, Kiranyaz S, Eren L, Askar M, Gabbouj M. Real-time motor fault detection by 1-D convolutional neural networks. IEEE Trans Ind Electron. 2016;63(11):7067–7075.
[15] Shao S, Wang P, Yan R. Generative adversarial networks for data augmentation in machine fault diagnosis. Comput Ind. 2019;106:85–93.
[16] Chen Z, Mauricio A, Li W, Gryllias K. A deep generative model for fault diagnosis of rolling element bearings. Mech Syst Signal Process. 2021;150:107335.
[17] Song Y, Gao S, Li Y, Jia L, Li Q, Pang F. Distributed attention-based temporal convolutional network for remaining useful life prediction. IEEE Internet Things J. 2021;8(12):9594–9602.
[18] Vaswani A, et al. Attention is all you need. In: Proc. NeurIPS; 2017.
[19] Nie Y, Nguyen NH, Sinthong P, Kalagnanam J. A time series is worth 64 words: Long-term forecasting with transformers. In: Proc. ICLR; 2023.
[20] Wu H, Hu T, Liu Y, Zhou H, Wang J, Long M. TimesNet: Temporal 2D-variation modeling for general time series analysis. In: Proc. ICLR; 2023.
[21] Lim B, Arık SO, Loeff N, Pfister T. Temporal fusion transformers for interpretable multi-horizon time series forecasting. Int J Forecast. 2021;37(4):1748–1764.
[22] Garza A, Mergenthaler-Canseco M. TimeGPT-1. arXiv:2310.03589 [Preprint]. 2023.
[23] Das A, et al. A decoder-only foundation model for time-series forecasting. In: Proc. ICML; 2024.
[24] Rasul K, et al. Lag-Llama: Towards foundation models for probabilistic time series forecasting. arXiv:2310.08278 [Preprint]. 2024.
[25] Yue Z, et al. TS2Vec: Towards universal representation of time series. In: Proc. AAAI; 2022.
[26] Chen T, Kornblith S, Norouzi M, Hinton G. A simple framework for contrastive learning of visual representations. In: Proc. ICML; 2020.
[27] Pan SJ, Yang Q. A survey on transfer learning. IEEE Trans Knowl Data Eng. 2010;22(10):1345–1359.
[28] Lu W, Liang P, Li Y, Liu L, Li M, Zhang J. Transfer learning for remaining useful life prediction of bearings under different working conditions. Rel Eng Syst Saf. 2022;225:108620.
[29] Raissi M, Perdikaris P, Karniadakis GE. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear PDEs. J Comput Phys. 2019;378:686–707.
[30] Karniadakis GE, Kevrekidis IG, Lu L, Perdikaris P, Wang S, Yang L. Physics-informed machine learning. Nat Rev Phys. 2021;3(6):422–440.
[31] Karpatne A, et al. Theory-guided data science: A new paradigm for scientific discovery from data. IEEE Trans Knowl Data Eng. 2017;29(10):2318–2331.
[32] Willard J, Jia X, Xu S, Steinbach M, Kumar V. Integrating scientific knowledge with machine learning for engineering and environmental systems. ACM Comput Surv. 2022;55(4):1–37.
[33] Grieves M. Digital twin: Manufacturing excellence through virtual factory replication. White paper; 2014.
[34] Tao F, Zhang H, Liu A, Nee AY. Digital twin in industry: State-of-the-art. IEEE Trans Ind Informat. 2019;15(4):2405–2415.
[35] Thelen A, et al. A comprehensive review of digital twin—part 1: Modeling and twinning enabling technologies. Struct Multidiscip Optim. 2022;65:354.
[36] VanDerHorn E, Mahadevan S. Digital twin: Generalization, characterization and implementation. Decis Support Syst. 2021;145:113524.
[37] Jones D, Snider C, Nassehi A, Yon J, Hicks B. Characterising the digital twin: A systematic literature review. CIRP J Manuf Sci Technol. 2020;29:36–52.
[38] Liu Y, et al. A novel cloud-based framework for the elderly healthcare services using digital twin. IEEE Access. 2019;7:49088–49101.
[39] Qi Q, et al. Enabling technologies and tools for digital twin. J Manuf Syst. 2021;58:3–21.
[40] Chao MA, Kulp C, Yeo CSK, Rocco CM, Zio E. Hybrid deep learning approach for RUL prediction. Rel Eng Syst Saf. 2022;222:108420.
[41] Yucesan E, Viana FA. A physics-informed neural network for wind turbine main bearing fatigue. Int J Prognostics Health Manage. 2020;11(1).
[42] Chen RTQ, Rubanova Y, Bettencourt J, Duvenaud DK. Neural ordinary differential equations. In: Proc. NeurIPS; 2018.
[43] Granger CWJ. Investigating causal relations by econometric models and cross-spectral methods. Econometrica. 1969;37(3):424–438.
[44] Spirtes P, Glymour CN, Scheines R. Causation, Prediction, and Search. Cambridge, MA, USA: MIT Press; 2000.
[45] Pearl J. Causality. Cambridge, U.K.: Cambridge Univ. Press; 2009.
[46] Runge J, Nowack P, Kretschmer M, Flaxman S, Sejdinovic D. Detecting and quantifying causal associations in large nonlinear time series datasets. Sci Adv. 2019;5(11):eaau4996.
[47] Lundberg SM, Lee SI. A unified approach to interpreting model predictions. In: Proc. NeurIPS; 2017.
[48] Ribeiro MT, Singh S, Guestrin C. ‘Why should I trust you?’: Explaining the predictions of any classifier. In: Proc. KDD; 2016.
[49] Wachter S, Mittelstadt B, Russell C. Counterfactual explanations without opening the black box: Automated decisions and the GDPR. Harv J Law Technol. 2017;31:841.
[50] Nauta M, Bucur D, Seifert C. Causal discovery with attention-based convolutional neural networks. Mach Learn Knowl Extr. 2019;1(1):312–340.
[51] Vovk V, Gammerman A, Shafer G. Algorithmic Learning in a Random World. Berlin, Germany: Springer; 2005.
[52] Romano Y, Patterson E, Candes E. Conformalized quantile regression. In: Proc. NeurIPS; 2019.
[53] Tibshirani RJ, Foygel Barber R, Candes E, Ramdas A. Conformal prediction under covariate shift. In: Proc. NeurIPS; 2019.
[54] Barber RF, Candes EJ, Ramdas A, Tibshirani RJ. Conformal prediction beyond exchangeability. Ann Statist. 2023;51(2):816–845.
[55] Gal Y, Ghahramani Z. Dropout as a Bayesian approximation: Representing model uncertainty in deep learning. In: Proc. ICML; 2016.
[56] Lakshminarayanan B, Pritzel A, Blundell C. Simple and scalable predictive uncertainty estimation using deep ensembles. In: Proc. NeurIPS; 2017.
[57] Abdar M, et al. A review of uncertainty quantification in deep learning: Techniques, applications and challenges. Inf Fusion. 2021;76:243–297.
[58] Peng W, Ye Y, Chen N. Bayesian deep learning for remaining useful life prediction of engineering systems. Rel Eng Syst Saf. 2019;191:106544.
[59] De Pater I, Mitici M. Conformal prediction for remaining useful life estimation of aircraft components. Rel Eng Syst Saf. 2023;233:109094.
[60] Geifman Y, El-Yaniv R. Selective classification for deep neural networks. In: Proc. NeurIPS; 2017.
[61] Madras D, Pitassi T, Zemel R. Predict responsibly: Improving fairness and accuracy by learning to defer. In: Proc. NeurIPS; 2018.
[62] Mozannar H, Sontag D. Consistent estimators for learning to defer to an expert. In: Proc. ICML; 2020.
[63] Lee JD, See KA. Trust in automation: Designing for appropriate reliance. Hum Factors. 2004;46(1):50–80.
[64] Amershi S, et al. Guidelines for human-AI interaction. In: Proc. CHI; 2019.
[65] Okamura K, Yamada S. Adaptive trust calibration for human-AI collaboration. Int J Hum-Comput Stud. 2023;172:102985.
[66] Ellefsen AL, Bjørlykhaug E, Æsøy V, Ushakov S, Zhang H. Remaining useful life predictions for turbofan engine degradation using semi-supervised deep architecture. Rel Eng Syst Saf. 2019;183:240–251.
[67] Mo Y, Wu Q, Li X, Huang B. Remaining useful life estimation via transformer-based architectures with multi-head attention. Rel Eng Syst Saf. 2021;210:107532.
[68] Tipaldi M, D’Arrigo PS, Macchia BC. Satellite anomaly detection using machine learning techniques. Adv Space Res. 2015;55(12):2884–2895.
[69] Ibrahim SK, Hameed MSA, Hameed KA. Machine learning approaches for spacecraft telemetry monitoring. In: Proc. IEEE Aerosp. Conf.; 2019. p. 1–10.
[70] Severson KA, et al. Data-driven prediction of battery cycle life before capacity degradation. Nat Energy. 2019;4:383–391.
[71] Stetco A, et al. Machine learning methods for wind turbine condition monitoring: A review. Renew Energy. 2019;133:620–638.
[72] Tautz-Weinert S, Watson SJ. Using SCADA data for wind turbine condition monitoring—A review. IET Renew Power Gener. 2017;11(4):382–394.
[73] Hu X, Xu L, Lin X, Pecht M. Battery lifecycle prognosis for electric vehicles. J Power Sources. 2020;448:227443.
[74] Fink O, Wang Q, Svense´n M, Dersin P, Lee WJ, Ducoffe M. Potential, challenges and future directions for deep learning in prognostics and health management applications. Eng Appl Artif Intell. 2020;92:103678.
[75] Zerveas G, Jayaraman S, Patel D, Bhamidipaty A, Eickhoff C. A transformer-based framework for multivariate time series representation learning. In: Proc. KDD; 2021.
How to cite this paper
@article{1723999,
author = {Atharv Atmaram Jadhav, Tushar D. Kolhe, Pradip M. Bhukan, Sakshi D. Dhembare, Bhagyashri R. Jadhav},
title = {Artificial Intelligence for Predictive Maintenance of Mission-Critical Engineered Systems: A Systematic Review of Methods, Frameworks, and Future Directions},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {10},
number = {4},
pages = {3011-3025},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1723999.pdf},
abstract = {Asset management in safety-critical industries is moving away from calendar-driven preventive schedules and toward autonomous, data-driven decision-making enabled by artificial intelligence (AI) and machine learning (ML). This work reports a PRISMA-compliant systematic review of the predictive maintenance (PdM) literature published between January 2018 and June 2026, covering domains such as aerospace, energy, and advanced manufacturing where component degradation must be tracked with high reliability. A structured search across four bibliographic databases (IEEE Xplore, Scopus, Web of Science, and the ACM Digital Library) produced an initial pool of 2,847 candidate records; after multi-stage screening, 127 peer-reviewed studies were retained for detailed synthesis. The retained literature is organized into six methodological threads: purely data-driven learning, physics-informed machine learning, digital-twin architectures, causal and explainable AI, uncertainty quantification, and human-AI collaboration. Although deep architectures achieve strong benchmark accuracy, the synthesis surfaces recurring shortcomings around out-of-distribution robustness, epistemic uncertainty handling, and model opacity. Eight recurring research gaps are catalogued, and a forward-looking agenda is proposed, culminating in a conceptual unifying framework, referred to here as NEXUS-AI, that combines physical priors, neuro-symbolic reasoning, and human oversight to support verifiable reliability in high-consequence settings.},
keywords = {predictive maintenance; prognostics and health management; remaining useful life; physics-informed machine learning; digital twin; explainable AI; causal inference; uncertainty quantification; systematic review},
month = {October},
doi = {https://doi.org/10.64388/IREV10I4-1723999}
}