Answers are grounded in this paper and cite their sources.
Agentic Artificial Intelligence (Agentic AI) represents a new class of autonomous systems capable of multi-step reasoning, tool orchestration, delegated action and adaptive interaction. Although these capabilities are gaining rapid traction in industry, their implications for maintenance management have not yet been systematically assessed. This paper conducts a process-centric literature review of Agentic AI using the eight-phase Maintenance Management Framework (MMF), mapping how agentic or precursor mechanisms are beginning to support diagnostic, predictive, optimization and supervisory functions across the maintenance lifecycle. A structured multi-database search identifies 42 relevant publications and reveals a marked imbalance: technical phases (diagnostics, Prognosis and Helath Management − PHM, scheduling and execution control) show empirical evidence, mainly driven by enabling precursors and a small number of agentic implementations, whereas managerial and strategic phases—asset criticality, KPI alignment, lifecycle costing and long-horizon decision-making—remain largely unexplored. Building on these findings, the paper proposes a six-stage workflow for the industrial introduction of Agentic AI, designed to prevent “pilot purgatory” and to ensure that autonomous capabilities evolve coherently from diagnostic assistance to strategic, lifecycle-oriented decision support. Each stage specifies the data, organizational and governance prerequisites required before progressing to higher autonomy and aligns the maturation path with the MMF phases. The review therefore delivers both a structured synthesis of the current state of Agentic AI in maintenance and a practical, implementation-oriented framework to guide its sustainable adoption in industrial environments.
No takes yet. Share an insight, caveat, or question.
Márquez et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: