What is reliability-centered maintenance?
Reliability-centered maintenance is a structured process for determining how physical assets should be managed so they continue to perform required functions in their current operating context. RCM evaluates functions, functional failures, failure modes, effects and consequences, then selects technically appropriate failure-management policies such as condition-based tasks, time- or cycle-directed tasks, failure-finding tasks, redesign or other actions, or deliberate run-to-failure where consequence and authority support it.
The discipline is the decision sequence, not the label.
A defensible RCM process asks what functions and performance standards matter, how those functions can fail, what causes each functional failure, what happens when the failure occurs, why the consequence matters, what proactive task can address it, and what should be done when no suitable proactive task exists.
Current SAE JA1011 provides evaluation criteria for processes represented as RCM. The value is in preserving that logic rather than shortening the process until every existing PM is simply relabeled RCM.
Different failure behaviors require different maintenance policies.
Condition-based tasks fit degradations that can be detected with enough warning and a reliable response path. Time- or cycle-directed restoration or replacement requires a defensible age relationship or governing requirement. Failure finding applies to hidden protective functions. Some failures are better managed through redesign, operating controls, or run-to-failure under controlled consequence.
The right answer can differ by failure mode on the same asset. Whole assets should not be forced into one strategy label.
Operating context determines whether a maintenance policy makes sense.
Redundancy, duty cycle, environment, production role, safety function, quality exposure, maintainability, spare availability, and failure detectability can change the decision. Copying a task from an identical model in different service can therefore reproduce the wrong strategy.
Consequence also changes the burden of proof. A low-consequence hidden function and a safety-protective hidden function do not deserve the same generic interval or test method.
RCM should evolve when operating evidence contradicts the original assumptions.
Work-order history, PM findings, condition data, failed tests, bad actors, recurrence, and RCA results all provide feedback. When a failure occurs despite the chosen policy, determine whether the failure mode was missing, the task was ineffective, the interval was wrong, execution failed, or the operating context changed.
RCM is therefore not a one-time workshop artifact. It is a governed maintenance strategy that should learn from the equipment.
What should an RCM analysis leave behind?
A traceable strategy for significant functions and failure modes, including consequence, selected failure-management policy, technical basis, governing authority, task or action requirements, ownership, and a review path when new operating evidence emerges.
- Start with function and operating context.
- Select policy at the failure-mode level.
- Use technical applicability before economics or convenience.
- Feed actual failures and task findings back into the strategy.
Technical references
- SAE JA1011_202411 — Current SAE standard defining evaluation criteria for RCM processes, revised November 2024.
- NASA Facilities Maintenance and Operations Management, Chapter 7 — NASA describes RCM as an ongoing process integrating proactive and reactive strategies and using operating data to improve maintenance and design.
Build maintenance strategy from function and failure behavior.
Use RCM logic to explain why a task exists, why its policy fits the failure, and what evidence will cause the strategy to be reviewed.
PM Optimization See facility-specific XRVE services