Most preventive maintenance programs are born the same way. Someone copies the intervals from the equipment manuals into a spreadsheet or a CMMS, the system starts generating work orders, and within six months the team is ignoring half of them because there are more PMs than hours. The program then gets blamed for not preventing anything.
The manual is the wrong starting point. A preventive maintenance program that holds up is built backward from failures and forward from the hours you have, and it covers fewer assets than you think.
What a preventive maintenance program is
A preventive maintenance program is a set of scheduled tasks, triggered by time, usage, or condition, designed to catch wear before it becomes a failure. Each task has an asset, a trigger, a procedure, and a record of completion. The program is the whole set plus the rules for adjusting it.
Preventive is not the same as planned or scheduled. Planned maintenance is any work prepared in advance. Scheduled maintenance is any work assigned to a date or meter reading. Preventive maintenance is the recurring subset intended to reduce failures. The terms overlap, and the confusion is why so many “PM programs” are really just calendars.
Why PM programs fail
The patterns repeat across plants and buildings.
Too many assets are in the program. A PM on a hand truck costs the same technician time as a PM on the main air compressor, and only one of them matters.
Intervals come from manufacturer manuals, which are written for warranty conditions, not for your duty cycle. A compressor running three shifts and a compressor running one shift get the same 90-day interval.
Time-based schedules sit on usage-driven wear. The forklift that ran 40 hours last month and the one that ran 400 get serviced on the same day.
Tasks say “inspect pump” with no checklist, no readings, and no pass or fail. Two technicians do two different inspections and neither one records anything useful.
Nobody checks completion, so overdue PMs pile up quietly. And when an asset fails anyway, nobody goes back to ask whether the PM should have caught it.
Seven steps to a program that holds
1. Decide which assets earn a PM. Rank your equipment by what happens when it fails. Production stops, safety risk, regulatory exposure, expensive repair, or none of the above. The top tier gets a full program. The middle tier gets a light one. The bottom tier gets run-to-failure, on purpose, in writing. Deciding to run something to failure is a maintenance strategy; forgetting to maintain it is an accident.
2. Start from your failure history. Pull two years of corrective work orders on the top-tier assets. For each failure, ask one question: what inspection or replacement, done earlier, would have caught this? Those answers are your first PM tasks. They will look nothing like the manual, and they will be right for your plant.
3. Pick the trigger that matches the wear. Calendar triggers fit age-driven wear: seals, belts, lubricant breakdown, filter loading in steady environments. Meter triggers fit usage-driven wear: compressor hours, vehicle mileage, press cycles, kilowatt-hours. Condition triggers fit anything you can measure cheaply, like vibration, temperature, or a differential pressure gauge on a filter. A rule that works: if usage varies by more than about a third between periods, use a meter, not a calendar.
4. Write tasks a new hire can complete without you. Every PM task gets a checklist with pass or fail items, required readings with acceptable ranges, and a place to note what was found. “Inspect belt” becomes “Check belt tension at midspan; acceptable deflection 10 to 15 mm; record measurement; replace if cracked or glazed.” The checklist is the procedure, the training document, and the record, all at once.
5. Set intervals you expect to change. Start from the manual for the first cycle. Then apply two rules. Tighten the interval on any asset that failed between PMs. Loosen it when five consecutive PMs find nothing to correct. Write the current interval and the reason for it in the task record. A year in, your intervals will be your own.
6. Protect PM hours on the calendar. A PM that always gets bumped for reactive work is a wish, not a program. Block PM hours per technician per week and treat them like production time. When reactive work overruns those hours three weeks in a row, that is a signal about staffing or about which assets belong in the program, and it deserves a meeting.
7. Measure four things and adjust quarterly. PM completion rate (done on time versus generated). Planned versus reactive hours as a percentage. Repeat failures on assets that are in the program. Findings per PM, meaning how often the checklist turns up something to fix. Low findings on an asset for a year means the interval is too tight or the asset does not belong. Review the four numbers every quarter and change intervals, tasks, or asset lists based on what they say.
Where software fits
A small team with a dozen critical assets can run this on a wall calendar and a binder of checklists. It breaks when meter readings need to trigger work, when the same task applies to twenty identical assets, or when overdue PMs need to escalate to someone before they become failures.
Preventive maintenance software inside a CMMS handles that mechanics: schedules by calendar, interval, or meter reading; work orders generated automatically when due; checklists with required readings attached to each task; one schedule applied across multiple assets; pending views and calendars for the planner; and notifications when a PM passes its threshold. MPACT CMMS by MPulse Software, as one example, lets the planner choose whether scheduled items convert to work orders automatically or wait for review, which matters when a plant shutdown moves everything by a week. Other CMMS platforms cover similar ground. The features that count are the meter trigger, the required fields on the checklist, and the escalation when something goes overdue.
Load only the top-tier assets first. A CMMS that generates two hundred PMs in its first month trains the team to ignore it.
A preventive maintenance checklist structure
Use this as the skeleton for every PM task:
- Asset and location (pre-filled from the schedule)
- Safety steps before starting (lockout, PPE, permits)
- Inspection items, each with pass or fail
- Readings, each with an acceptable range
- Replacements or adjustments made
- Findings that need a follow-up work order
- Time on task and technician name
Print it or put it on a phone. Either way, the technician fills in every line, and the readings are numbers, not words.
The honest summary
A preventive maintenance program is a bet that a small amount of scheduled work costs less than the failures it prevents. The bet pays off only on assets whose failures matter, with tasks specific enough to find something, at intervals tuned to how the equipment is used. Start with your worst failures, cover your most critical assets, protect the hours, and revise every quarter. The software makes the schedule reliable. The judgment about what to schedule stays with you.
