Digitalization in Industrial Facilities and the Classification of Maintenance Types: A Review
Abstract
The digitalization of industrial facilities, driven by Industry 4.0 technologies, is fundamentally changing how production assets are monitored, modeled, and managed, and its strongest effect is felt in maintenance management, where traditional corrective and time-based practices are giving way to data-driven, condition-based, and predictive maintenance strategies. This review addresses two linked questions: how industrial maintenance types can be organized within a consistent framework, and how digital enablers such as cyber-physical systems, the Industrial Internet of Things (IIoT), the digital twin, machine learning, and remaining-useful-life (RUL) estimation reshape these types. The study is designed as a narrative (non-systematic) review of seventy-one peer-reviewed journal articles published between 1995 and 2026, complemented by the primary sources of the reliability centered maintenance literature. Maintenance strategies are classified and compared along a reactiveproactive axis as corrective, planned/periodic, condition-based, predictive, reliability-centered (RCM), risk-based (RBM), and opportunistic maintenance. We show that digitalization repositions these types within a continuous condition-monitoring–diagnosis–prognosis–decision loop. The review also surveys recent developments (large language models and generative artificial intelligence, federated and edge learning, explainable AI, digital-twin interoperability standards, and the human-centric maintenance perspective of Industry 4.0) and discusses their implications for maintenance management. As an original contribution, we propose a conceptual digital-maintenance framework that couples the digital twin, RUL-based maintenance-interval computation, and an expected profit-loss indicator (ePLI) with a reminder/alarm decision layer; the ePLI is formalized through an expected-loss expression, a threshold-based reminder/alarm rule, and an illustrative application scenario. The findings indicate that digitalization turns maintenance types from mutually exclusive choices into elements of an integrated asset-management continuum, and that Industry 4.0 and trustworthy-AI research adds a human-centric dimension to this continuum. For practitioners, the r