How do the seven RCM questions lead to a maintenance strategy?
Reliability-Centered Maintenance evaluates required asset functions and performance standards, functional failures, causes or failure modes, effects, consequences, applicable proactive tasks, and the default actions when no suitable proactive task exists. The resulting policy may be condition-based, time- or cycle-directed, failure-finding, redesign, operating controls, or intentional corrective maintenance, subject to governing safety and technical requirements.
Ask the seven RCM questions in sequence and keep the answer traceable.
SAE JA1011_202411 defines criteria for processes to be evaluated as RCM. A workshop that simply lists existing PMs and calls them 'RCM tasks' has not answered the functional and consequence questions.
| Question | Output the team should record |
|---|---|
| 1. What must the asset do? | Function, performance standard, duty, operating context. |
| 2. How can it fail functionally? | Loss, partial loss, or unacceptable performance. |
| 3. What causes each functional failure? | Credible failure modes at an actionable level. |
| 4. What happens when each mode occurs? | Observable local and system failure effects. |
| 5. Why does each mode matter? | Safety, environment, operational, economic or other consequences. |
| 6. What proactive task is applicable and effective? | Technically justified task, trigger, capabilities and feasibility. |
| 7. What if no suitable proactive task exists? | Controlled default action: failure-finding, redesign, policy change, or corrective maintenance as applicable. |
Select a policy for each failure mode, not a label for the whole asset.
| Failure behavior | Candidate policy | Key technical question |
|---|---|---|
| Degradation detectable before functional failure | Condition-based inspection or predictive monitoring | Is the potential failure detectable with enough reliable warning and a response process? |
| Clear age or usage relationship | Time- or cycle-directed restoration/replacement | Does age-based intervention reduce the consequence in a defensible way? |
| Hidden protective function could fail without indication | Failure-finding test under authorized engineering basis | Can the hidden failure be revealed safely, with test coverage and frequency justified? |
| No viable task or unacceptable consequence | Redesign, operational control or other authorized action | Can the consequence be reduced acceptably by changing the system or process? |
| Tolerable evident non-safety consequence | Controlled corrective maintenance or run-to-failure | Is consequence accepted, restoration feasible and the decision documented? |
One pumping system can legitimately have several maintenance policies.
Consider a fictional utility-pumping skid. A critical duty pump bearing may justify condition monitoring if a validated method detects deterioration in time to intervene. A standby pump's hidden auto-start function may justify authorized functional testing. A noncritical status lamp might be corrected on failure where alternate indications, operating requirements, and consequence allow it.
All three belong to one asset system, but none justifies assigning one blanket PM frequency to the whole system. The correct decision depends on specific failure modes, required function, detection capability, and consequence.
A PM task must be technically applicable before its frequency is discussed.
Ask whether the method can reveal the failure mode, what threshold constitutes an actionable finding, whether the P-F interval is known with adequate confidence, and whether planners can convert findings into timely work. For replacement or restoration, verify age dependence or other controlling requirements before selecting a calendar interval.
A failure-finding task requires special care because the failure is hidden; generic proof-test steps and test intervals cannot be authorized by an article. OEM, engineering, safety, regulatory, and site procedures remain controlling.
RCM is a living decision record, not a one-time workshop output.
Store function, mode, consequence, selected task or default action, decision logic, governing authority, implementation owner, and confidence. Keep changes connected to asset hierarchy and CMMS jobs so technicians can execute what the analysis recommends.
Review when the operating context changes, a high-consequence event occurs, PMs repeatedly find nothing, a protective test fails, or bad-actor and RCA evidence contradicts the strategy. Avoid treating 'no new failure in 30 days' as proof of effectiveness when exposure is too limited to evaluate the risk.
What should an RCM facilitator refuse to sign off?
A policy with no identified failure mode, no clearly stated consequence, no technical applicability argument, no implementation owner, or an interval that conflicts with controlling OEM, engineering, or regulatory requirements.
- Use RCM at the failure-mode level.
- Make safety and environmental consequences explicit.
- Use FMEA to support but not replace the RCM decision process.
- Retain uncertainty, authority and review triggers.
Technical references
- SAE JA1011_202411, Evaluation Criteria for RCM — Current RCM process-evaluation standard, revised November 2024.
- NASA Reliability-Centered Maintenance Guide — Connects failure modes, effects, consequences, and maintenance policy selection.
Leave with a policy that can be executed and challenged.
Record the function, mode, consequence, technically applicable policy, controlled task or redesign action, and evidence that will trigger a later review.
RCM learning path See facility-specific XRVE services