Predictive Maintenance and Condition Monitoring in Critical Power and Energy Infrastructure

@article{1722489,
      author = {Tunji S. Adaramola, Samuel Fadero, Edikan Nse Gideon},
      title = {Predictive Maintenance and Condition Monitoring in Critical Power and Energy Infrastructure},
      journal = {Iconic Research And Engineering Journals},
      year = {2018},
      volume = {2},
      number = {5},
      pages = {454-477},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1722489.pdf},
      abstract = {This review examines the development, technologies, applications, and implementation requirements of predictive maintenance and condition monitoring across critical power and energy infrastructure. Its purpose is to determine how contemporary asset-management approaches can improve reliability, safety, operational continuity, lifecycle value, and resilience in increasingly complex energy systems. The study adopts a structured narrative review of established scholarly literature published up to 2018, synthesising evidence on maintenance philosophies, degradation mechanisms, sensing technologies, data architectures, diagnostic analytics, machine learning, digital twins, remaining-useful-life estimation, cybersecurity, and economic evaluation. The findings show a decisive transition from reactive and interval-based maintenance towards condition-based, predictive, and risk-informed decision-making. Effective implementation depends on the coordinated use of vibration, thermal, electrical, chemical, acoustic, and operational data, supported by interoperable communication networks, secure storage, rigorous data-quality management, signal processing, statistical analysis, and physics-based models. Machine learning and digital twins strengthen fault detection and prognostics, although their value is constrained by scarce failure data, model uncertainty, legacy-system incompatibility, cybersecurity exposure, workforce limitations, and weak economic justification. Applications across conventional generation, transmission, distribution, renewable-energy, and storage assets demonstrate that predictive methods can reduce unplanned outages, improve maintenance prioritisation, extend asset life, and enhance service resilience. The review concludes that predictive asset management should be treated as an integrated organisational capability rather than a stand-alone technological intervention. It recommends phased deployment focused on high-criticality assets, stronger data governance, validated and explainable models, secure interoperable architectures, workforce development, and lifecycle-based investment appraisal. Future research should prioritise affordable edge analytics, uncertainty-aware prognostics, climate-responsive degradation models, standardised datasets, and cybersecure digital-twin frameworks suited to diverse operating contexts.},
      keywords = {predictive maintenance, condition monitoring, critical infrastructure, asset reliability, digital twins, operational resilience.},
      month = {November},
  }