XRVE Reliability knowledge base | Preventive maintenance

A good PM task tells the technician what failure evidence to look for and what happens when it is found.

Preventive maintenance becomes weak when tasks describe activity without a technical purpose. A defensible task connects the asset function and failure mode to observable evidence, a controlled method, a decision boundary from the proper authority, and a defined response when the condition is unacceptable.

Published October 4, 2026 Reviewed and authored by Joshua Rivera Maintenance & Reliability Leader

How should a preventive maintenance task be written?

A preventive maintenance task should state the equipment or function being protected, the failure mode or degradation mechanism the task addresses, what evidence the technician must inspect or act on, the approved method or reference that governs the work, the acceptance criteria or decision boundary, and what corrective action or work request should be initiated when the condition does not meet the requirement.

A task is not technically strong merely because it contains a verb.

Instructions such as inspect, check, lubricate, clean, or verify can describe work without explaining what degradation the work is intended to prevent or detect. When the task purpose is invisible, technicians cannot distinguish useful evidence from routine completion.

The task-writing standard should make the technical intent traceable enough that planners, technicians, reliability staff, and future reviewers can understand why the task exists.

XRVE field note

If a technician can complete the PM perfectly and still have no way to tell whether the targeted failure mechanism is developing, the task may be compliant but ineffective.

Build each PM task around an evidence chain.

A useful task normally identifies the protected component or function, credible failure mode, observable condition, controlled method, decision criterion, and follow-up path. The CMMS wording does not need to duplicate an approved procedure; it should point to the controlling source when that source governs technical details.

Task elementQuestion it should answer
Protected function or componentWhat are we trying to keep capable?
Failure modeWhat specifically can go wrong?
EvidenceWhat can the technician observe, measure, detect, or confirm?
Method/referenceWhich approved source governs the work?
Acceptance criterionWhat condition separates acceptable from follow-up?
Corrective triggerWhat happens when the condition is unacceptable?

Do not invent technical limits merely to make the PM look complete.

Torque, vibration, temperature, alignment, pressure, wear, clearance, lubrication, electrical, safety-device, and quality limits may require OEM documentation, engineering authority, approved site standards, regulatory requirements, or another controlled source. A content standard should never fabricate those values.

Where a numerical limit is not defensibly available, identify the required authoritative reference, capture the observed condition, and route the finding for qualified review rather than substituting an unsupported threshold.

A finding without a response path is incomplete preventive maintenance.

The PM should make it clear how an unacceptable condition becomes controlled corrective work. That may be a linked notification, work request, follow-up order, immediate escalation under site rules, or another governed workflow.

Without that loop, the organization can collect defects repeatedly while still reporting high PM completion. Effectiveness depends on whether useful findings lead to appropriate action before functional failure.

Review the task for technical purpose, execution clarity, and feedback value.

Ask whether the failure mechanism is credible, whether the task can detect or prevent it at a useful point, whether the technician has a defensible method, whether the decision boundary comes from proper authority, and whether the result becomes usable maintenance history.

Tasks that cannot satisfy those questions should be revised, consolidated, replaced by a different strategy, or removed when evidence no longer supports them and the required authority approves the change.

What should happen when a PM task cannot explain its technical purpose?

Treat the task as a review candidate rather than automatically preserving it because it has always existed. Verify the function, failure mode, evidence, method, decision criterion, and downstream response before deciding whether the task should remain.

  • Tie the task to a credible failure mechanism or protected function.
  • Use controlled technical references for limits and methods that require authority.
  • Define what evidence the technician should capture.
  • Create a clear finding-to-corrective path.
Use the method, then use the evidence

Make every PM task earn its place in the program.

A strong PM library is not the largest one. It is a controlled set of tasks whose technical purpose, evidence, response path, and maintenance history can be defended.

PM Optimization See facility-specific XRVE services