Meter-based (usage-based) PM triggers maintenance on how much an asset has run, operating hours, cycles, units, or miles, instead of on the calendar. A forklift is serviced every 250 engine hours, not every month, so heavily used machines get more care and idle ones are not over-serviced.
The calendar is a lazy proxy for wear. A pump that ran flat out all month and a pump that sat idle both trip the same "monthly PM," even though only one earned it. Meter-based PM fixes that by tying the trigger to the thing that actually causes wear: usage. It sits between blunt time-based PM and full condition-based maintenance and for a huge class of equipment it is the most cost-effective trigger you can run. This guide covers what it is, when to use it, the meters that matter, where it goes wrong, and how to set it up.
What is meter-based PM?
Meter-based PM is preventive maintenance scheduled against a usage meter rather than a date. You define a threshold, every 250 operating hours, every 100,000 cycles, every 5,000 miles, and when the asset's meter reaches it, the PM work order fires. The reading comes from somewhere: a technician logging a counter on a route, or telematics and connected controls pushing the value automatically. Either way, the job is driven by what the machine did, not by what day it is.
The payoff is that maintenance effort follows the wear. Two identical machines with very different duty cycles get very different service frequencies, automatically, without anyone deciding it month by month. That both prevents under-maintenance (a hard-run machine blowing past its safe interval on a fixed monthly plan) and over-maintenance (tearing into a lightly used machine that did not need it, and risking the infant-mortality failures that intrusive work introduces).
How is meter-based PM different from calendar-based PM?
The difference is the trigger. Calendar-based PM counts days; meter-based PM counts usage. On steady, predictable duty they converge, if a machine runs the same hours every week, "every 30 days" and "every 250 hours" land in about the same place. They diverge exactly when duty is variable, which is where meter-based earns its keep.
| Calendar-based PM | Meter-based PM | |
|---|---|---|
| Trigger | Elapsed time (every 30 days) | Accumulated usage (every 250 hours) |
| Best when | Wear tracks time; duty is steady | Wear tracks usage; duty varies |
| Failure mode | Over- or under-services variable-duty assets | Misses time-driven decay on idle assets |
| Data needed | Just a date | A reliable meter reading |
| Good for | Inspections, calibration, seasonal tasks | Engines, compressors, vehicles, presses |
When should you use meter-based instead of time-based triggers?
Use a meter when the asset's wear is driven by how much it runs, not how long it exists. The clearest signals: the equipment has variable or unpredictable duty, its failure modes are usage-driven (bearing wear, oil degradation, cycle fatigue), and it already has a trustworthy counter. Engines, compressors, vehicles, pumps, and cyclic machines like presses and packaging equipment are the classic fits, an oil change belongs at 250 running hours whether that took three weeks or three months.
Stay on the calendar when decay is time-driven regardless of use: rubber and elastomers aging, lubricant oxidizing, corrosion, and regulatory or safety tasks like lubricant condition checks calibration, and inspections that must happen on a fixed cadence no matter what the meter says. And reach past both toward condition-based or predictive triggers when you can actually measure the degradation directly, vibration, temperature, oil analysis, because measuring the condition beats estimating it from a usage proxy.
What meters do plants actually track?
Anything that counts up with use can be a PM meter. The common ones: operating hours (engines, compressors, motors, gensets), cycles or strokes (presses, valves, actuators, molding machines), units produced (fillers, packaging lines, tie the PM to output), distance (fleet vehicles, forklifts, conveyors), and process-specific counters like gallons pumped, tons processed, or rounds fired. The trick is picking the meter that best correlates with the wear you are trying to prevent. For a hydraulic press, cycles predict wear better than hours, because a press idling under power ages its rams far slower than one stamping parts. Choose the meter that moves when the damage happens.
Where does meter-based PM go wrong?
Bad or missing readings. The whole scheme depends on the meter value being current and correct. Manual logs get skipped, transposed, or guessed; a meter that is read late fires the PM late. Automated capture from telematics or connected controls removes this failure mode, which is the same argument as for trustworthy MTBF data and any other machine-monitored metric, the reading has to survive an audit.
Ignoring time-driven decay. A pure meter trigger never services a machine that sits idle for a year, even though its seals dried out and its oil oxidized. The standard fix is a "whichever comes first" rule: 250 hours or 6 months, whichever arrives sooner. This combined trigger catches both usage wear and calendar decay, and it is the setup most mature programs actually run.
Meter rollover and resets. Counters roll over, get replaced, or reset to zero, and a naive system reads that as negative usage or a giant jump. Handle rollover explicitly, and record meter swaps, or you will get phantom PMs or missed ones.
Fixed versus floating schedules. Decide whether the next PM counts from the interval mark or from when the last one was actually done. Floating from actual completion prevents a backlog of "catch-up" PMs stacking when one runs late, usually the healthier choice, and a detail your planning and scheduling process should settle deliberately.
How do you set up a meter-based PM program?
Rolling out meter-based triggers is a short project with a clear order of operations. The core loop is simple: a reading comes in, the system compares it to the threshold, and if the threshold is crossed a work order is born, no one has to watch a counter.
- Find the usage-driven assets. Screen your critical equipment for wear that tracks usage rather than time, engines, compressors, cyclic machines, fleet. These are your candidates; leave time-driven tasks on the calendar.
- Pick the meter that matches the wear. Hours, cycles, units, or distance, choose the one that moves when the damage happens. The wrong meter gives you a precisely wrong schedule.
- Establish a reliable reading source. Automate it from telematics, a PLC, or connected controls where you can; where you cannot, put meter reads on a technician route with a clear owner. No reading, no trigger.
- Set the interval from real data. Base thresholds on manufacturer guidance, then tune with your own failure history and oil-analysis results. A generic 250 hours is a starting point, not a law.
- Add a "whichever comes first" calendar backstop. Pair each meter interval with a maximum time interval so idle assets still get their time-driven service. Decide fixed versus floating scheduling here too.
- Automate the work order. Let the system raise the PM automatically when the threshold is crossed, so no one is watching counters by hand, this is core CMMS functionality, and it feeds the same work order workflow as every other job.
- Review and re-tune. Watch whether failures cluster before the interval (tighten it) or PMs keep finding nothing wrong (loosen it), and track the effect on MTBF and MTBR. The interval is a hypothesis you keep testing.
What the numbers say
- Right-timing PM against usage is squarely where maintenance savings live. The U.S. Department of Energy's FEMP O&M guidance, maintained by PNNL, finds that moving beyond blunt time-based work toward condition- and usage-driven maintenance saves 8–12% over a purely preventive (time-based) program and the opportunity versus reactive, run-to-failure operation can exceed 30–40% (PNNL, O&M Best Practices: Maintenance Approaches). Meter-based triggers capture much of that gain without the instrumentation cost of full condition monitoring.
- Every PM that fires too early also burns scarce technician time. The U.S. Bureau of Labor Statistics projects 13% employment growth (2024–2034) for industrial machinery mechanics, machinery maintenance workers, and millwrights, much faster than average, with about 538,300 jobs in 2024 and roughly 54,200 openings a year (BLS Occupational Outlook Handbook). Not over-servicing idle equipment is a direct way to spend that scarce labor where it earns.
Meter-based PM is a simple idea with a big return: stop pretending the calendar knows how hard a machine worked. Trigger on usage, back it up with a time limit for idle decay, feed it reliable readings, and let the system raise the work order, and your maintenance effort finally lands where the wear actually is. It is one of the highest-return upgrades on the path from reactive firefighting to genuine equipment reliability. For how one manufacturer built floor data trustworthy enough to trigger maintenance on real machine signals, see the CLS case study or see how the pieces connect on the Harmony platform.