PM Optimization Using Failure Data

Facility Maintenance
1

Why This Topic?

Many PM programs are built on:

  • OEM recommendations

  • Legacy practices (“this is how it’s always been”)

  • Copy-paste CMMS setups

Yet breakdowns still happen—and PMs keep piling up.

Honest question: Are your PMs actually preventing failures… or just filling calendars?

The Core Idea

Failure data should drive PM optimization.

Every corrective work order tells a story:

  • What failed

  • How often it failed

  • How long it took to fix

  • What it cost (labor, parts, downtime)

Ignoring this data turns PMs into assumptions instead of controls.

Common Issues Seen in Facilities

  • PMs added after every failure without analysis

  • Repeat failures on the same asset

  • PM intervals that don’t match real-world usage

  • PM tasks that don’t address the actual failure mode

  • No review cycle for PM effectiveness

If this sounds familiar, you’re not alone.

What Failure Data Is Actually Useful?

Instead of focusing on single incidents, look for patterns:

  • Repeat corrective WOs on the same asset

  • Same failure mode occurring multiple times

  • MTBF trending downward

  • Failures occurring shortly after PMs

  • High-cost or high-downtime failures

These usually point to PM design gaps—not technician performance.

What PM Optimization Looks Like in Practice

Facilities that optimize PMs typically:

  • Adjust PM frequency based on failure trends

  • Modify PM steps to target known failure modes

  • Add condition checks instead of adding more PMs

  • Eliminate PMs that don’t reduce failures

  • Validate changes using MTBF and reactive vs planned ratios

Optimization is about precision, not cutting corners.

Common Mistakes to Avoid

  • Making PMs longer instead of more focused

  • Overreacting to one failure

  • Ignoring environmental and usage factors

  • Changing PMs without documenting why

  • Never reviewing the impact of changes

Field Reality Check

  • More PMs ≠ better reliability

  • Failure history beats OEM intervals

  • One targeted PM can replace several generic ones

  • If techs consistently question a PM, it deserves review

Let’s Discuss

  • Which PM in your facility adds the least value?

  • Have you ever reduced or removed a PM? What was the outcome?

  • Do your PM tasks align with real failure causes?

  • How often do you review PM effectiveness using actual data?

1 Like