How should a plant decide which preventive maintenance tasks to improve or remove?
Evaluate every PM task against the failure mode it is meant to prevent or detect. A completed checklist is not proof of risk reduction. Optimize the strategy, not just the number of work orders.
Why PM libraries need regular review
PM libraries tend to grow after failures, vendor recommendations and audits. New tasks are added but older ones are rarely challenged. Plants can therefore have high PM compliance and still experience repeated failures. The correct question is not whether a task was completed, but whether it controls a meaningful failure mechanism and produces an actionable result.
Optimization should consider safety, environmental and production consequences, labor, intrusive maintenance risk and the cost of failure. Reducing work without evaluating those factors is not reliability improvement.
Map tasks to failure modes
For each asset, define its required function, how that function can be lost, and the credible causes. Review work order history, operating duty, inspection findings and manufacturer requirements. Map each existing PM task to one or more failure modes and record the intended effect.
- Condition based inspection: Detect deterioration early enough for planned action.
- Scheduled replacement: Address a failure mode with evidence of age or usage dependence.
- Failure finding: Test a hidden protective function under an approved procedure.
- Design or operating change: Address a cause that routine maintenance cannot control effectively.
- Run to failure: Use deliberately only where consequences and recovery arrangements justify it.
Ask whether each task works
Can the technician explain the defect being sought? Is there a measurable acceptance criterion? Is the interval supported by deterioration behavior or a mandatory requirement? What happens when an abnormal condition is found? If no one can answer, mark the task for review rather than silently deleting it. Generic instructions such as “inspect pump” rarely create repeatable evidence.
Example and pilot approach
Suppose a pump passes a monthly visual PM yet suffers recurring bearing damage. The team compares the failure evidence with the inspection instructions. If contamination or alignment is the credible cause, the team may need improved contamination control, more specific measurements or a suitable monitoring method rather than another generic check. This example is illustrative.
- Select a manageable group of important assets.
- Classify tasks as retain, clarify, change interval, replace, or evaluate for removal.
- Obtain approval from operations, reliability, safety and the appropriate technical authority.
- Update CMMS instructions, alert limits and corrective work routing.
- Track repeat failures, useful findings, emergency work and labor after the change.
Safety and governance limits
Do not remove legally required inspections, protective device tests, manufacturer required tasks or warranty obligations through a cost cutting exercise. Changes to safety critical functions may require formal technical review and management of change. Preserve the rationale for every change so future teams understand why a task exists or was retired.
What success looks like
A good program may contain fewer tasks but produce more useful findings and fewer repeat failures. Each task has a clear purpose, acceptance criterion, owner and response path. Measure failure prevention and quality of findings alongside completion percentage.
Explore the assessment
KANE can help evaluate the underlying maintenance process and build a practical improvement plan.
View the related assessment →Related answer: Why do equipment failures happen even when preventive maintenance is completed?