Preventive maintenance is supposed to stop failures before they happen. The trick is figuring out when to do it. Too often, you waste labour. Too rarely, you lose assets. The right answer depends on the asset, the failure mode, and what you can measure. Three strategies cover 95% of the cases.
The three strategies
Time-Based
Fixed calendar intervals. Simplest to plan. Risks over- or under-servicing.
Meter-Based
Triggered by usage: runtime hours, cycles, distance. Matches wear to actual operation.
Condition-Based
Triggered by measured asset condition (vibration, temperature, oil analysis). Most efficient, most complex.
Time-Based Maintenance
Scheduled at fixed intervals regardless of actual usage. Monthly, quarterly, annually. The default for most PMs because it's simple to plan, resource, and audit.
When to use:
- Assets with consistent, predictable usage
- Regulatory/compliance inspections (fire, safety)
- Low-criticality assets where over-maintenance is cheap
- When you can't meter usage easily
Downsides:
- Over-maintains assets that aren't heavily used
- Under-maintains assets with high usage variance
Meter-Based Maintenance
Triggered when an asset hits a usage threshold: runtime hours, number of starts, cycles, distance, throughput. Matches PM frequency to actual workload.
When to use:
- Assets with variable duty cycles (backup generators, peak-load equipment)
- Manufacturer-specified service intervals ("change oil every 500 hours")
- Wear-driven components where failure correlates with use
- Mobile fleet (distance-based)
Downsides:
- Requires reliable meter readings (manual or automated)
- Less predictable resource planning
Condition-Based Maintenance
Triggered when measured condition crosses a threshold. Vibration analysis, thermal imaging, oil particle counts, ultrasonic testing. Most efficient because you only maintain what needs it, when it needs it.
When to use:
- Critical assets where failure is expensive
- Assets where failure modes are detectable in advance
- Operations with IoT sensors or predictive analytics platforms
- Where you have skilled reliability engineers to interpret data
Downsides:
- Requires instrumentation (sensors, data platforms)
- Requires skilled analysts
- Higher setup cost; pays off on critical assets only
Combining strategies on one asset
Most critical assets use all three
A chiller might have: annual inspection (time-based, regulatory), oil change every 2000 run-hours (meter-based), and vibration monitoring with threshold alerts (condition-based). Each strategy covers a different failure mode at the appropriate frequency.
How to select a strategy
- Start with asset criticality. Critical assets justify condition-based investment.
- Check manufacturer recommendations. Some schedules are non-negotiable.
- Assess what you can reliably meter or measure.
- Consider labour availability. Condition-based needs analysts, not just technicians.
- Default to time-based for everything else.
Common strategy pitfalls
- Over-applying time-based. Every asset on quarterly inspection regardless of use.
- Meter-based without meter readings. Strategy set up, readings never captured. Schedules drift.
- Jumping to condition-based too early. Before you have asset criticality data or trained staff.
- Not reviewing schedules. Set once, never tuned. Failure data should feed back into frequency adjustments.
Conclusion
PM strategy is about fit, not sophistication. Time-based is fine for 70% of assets. Meter-based adds value where usage varies. Condition-based is reserved for the critical 5-10% where prevention truly pays. Match the strategy to the asset and don't over-engineer your PM program.
Written by Muhammad Abbas
Enterprise integration specialist designing PM frameworks across EAM and CMMS platforms.
