Preventive maintenance
Preventive maintenance: types, real examples and how to start in one week
Preventive maintenance is planned work you do on a machine before it fails: cleaning, lubrication, inspection, adjustment and replacing parts that wear, on a schedule set by time, usage or condition. You spend a little planned time to lose less unplanned time. You can start in one week with your most critical machines, a short checklist for each, and one person who owns the schedule.
What is preventive maintenance?
Preventive maintenance (PM) is maintenance you plan and carry out before a failure, to keep a machine able to do its job. The work is usually small: clean, lubricate, inspect, tighten, adjust, and replace parts that wear on a known pattern, such as filters, belts, hoses and seals.
It is different from breakdown maintenance, where you repair after the machine stops, and from corrective work raised when an inspection finds a defect. A working PM programme turns some of tomorrow's breakdowns into today's short jobs, done in a planned stop instead of in the middle of an order.
PM is not the right answer for every asset. A cheap, non-critical item with a spare on the shelf, such as a pedestal fan, can run to failure by choice. The point is to decide machine by machine, so the default is no longer "fix it when it breaks".
In plants that follow total productive maintenance (TPM), operators take on the simplest PM tasks, such as daily cleaning, lubrication and visual checks, and the maintenance team handles the rest. See our guide to total productive maintenance for how that split works.
The four types, and when to use each
Most plants use a mix. Pick the trigger that matches how the machine actually wears, not the one that is easiest to put in a spreadsheet.
| Type | What triggers the work | Example | Good fit for |
|---|---|---|---|
| Time-based | A calendar interval: daily, weekly, monthly, yearly | Monthly cleaning of the filters on a control panel | Assets that wear steadily, and statutory or safety checks |
| Usage-based | Run hours, cycles, strokes or output since the last PM | Compressor oil and filter change after a set number of running hours | Machines whose load changes with the production plan |
| Condition-based | A reading crosses a limit: vibration, temperature, pressure | Bearing inspection when pump vibration passes its warning limit | Critical rotating equipment where a reading shows wear early |
| Predictive | A trend in condition readings shows when a limit will be crossed | Planning a motor bearing change for the next planned stop from a rising temperature trend | High-value assets with reliable readings |
A useful fifth option is "whichever comes first": the PM falls due on a date or at a usage interval, whichever is reached first. It suits machines that sit idle for weeks and then run hard.
Real examples from common plant machines
Copy this as a starting checklist. The intervals are typical starting points, not rules: your OEM manual, your operating hours and your own breakdown history decide the final interval.
| Machine | PM task | Trigger | Starting interval |
|---|---|---|---|
| Screw air compressor | Check oil level, drain condensate, clean the intake filter | Time | Daily check, weekly filter clean |
| Screw air compressor | Change oil, oil filter and air filter element | Usage | OEM running-hour interval |
| Hydraulic press | Check oil level and temperature, look for leaks at hoses and fittings | Time | Every shift |
| Hydraulic press | Replace the return-line filter, take an oil sample | Whichever comes first | OEM hours or six months |
| CNC machining centre | Clean the chip conveyor and coolant tank, check way lube level | Time | Daily and weekly |
| CNC machining centre | Record spindle temperature and vibration against a baseline | Condition | Monthly reading |
| Centrifugal pump | Check the seal for leaks, bearing temperature and vibration | Condition | Weekly round |
| Belt conveyor | Check belt tension, tracking and idlers, grease the bearings | Time | Weekly |
| EOT crane or hoist | Inspect wire rope, hook, brake and limit switches | Time | Monthly in-house check, plus the statutory examination your factory rules require |
Every task needs three things written down: what to check, the acceptable limit, and what to do when the reading is outside it.
How to start in one week
You do not need software, a consultant or a full asset register to begin. You need one owner, a short list of machines and the discipline to record what you find. Here is a plan that fits in one working week.
- Day 1: list and rank. Walk the floor and list your machines. Mark the ones that stop the line, are hard to repair or have no standby. Pick about ten.
- Day 2: gather what you know. Collect the OEM manuals and the last year of breakdowns for those machines, from registers, WhatsApp messages and the shift log. Note what failed and how often.
- Day 3: write short checklists. For each critical machine, write five to ten tasks from the manual and the failure history. Plain words, one check per line, with the limit.
- Day 4: set triggers. Give each task a trigger and an interval: calendar, run hours, or whichever comes first. Where you have no data, start with the OEM interval.
- Day 5: assign owners and slots. Name who does each task, operator or technician. Agree with production when the planned stop is, so PMs do not compete with orders.
- Day 6: run the first PMs. Do the first round, record readings and photos, and raise a work order for every defect found.
- Day 7: review. Count what was done, what was missed and why, and what the PMs found. Fix the checklists that were unclear.
After the first week, keep the same ten machines for a month before you add more. A small plan that runs every week beats a large plan that runs once.
A worked example from a typical Indian plant
Worked example (illustrative)
Take a forging and machining plant near Pune with about 140 assets. In week one the team picks 12 critical machines: two compressors, four hydraulic presses, four CNCs, a cooling-water pump and an EOT crane. They write 38 PM tasks across those machines. All numbers below are illustrative.
PM compliance is the share of PMs done within their window: PMs done on time divided by PMs due in the period, times 100. In the first month, 64 PM jobs fall due and 51 are done on time. Of the 13 missed, 9 were skipped because the machine was needed for an urgent order.
| Month 1 (illustrative) | Calculation | Result |
|---|---|---|
| PM jobs due | From 38 tasks on 12 machines | 64 |
| Done within the window | 51 | |
| PM compliance | 51 / 64 x 100 | 79.7% |
| Missed because of a production clash | 9 of 13 | |
| One press hose failure, cost of the stop | 5.5 h x ₹9,000 lost output per hour + ₹6,000 hose, oil and clean-up | ₹55,500 |
What to do with it. The 9 clashes are a planning problem, not a maintenance one: move those PMs into the weekly planned stop and agree that slot with production. The hose failure shows why a 20-minute weekly hose and fitting check on each press earns its place. Track PM compliance and breakdown hours on the same 12 machines for three months before you widen the programme.
Common mistakes
- Copying the whole OEM manual on day one. Two hundred tasks nobody can finish get ignored. Start with critical machines and the tasks that match their failures.
- Setting every interval by the calendar. A press on three shifts wears faster than the same press on one. Use run hours, or whichever comes first.
- Checklists that say "check machine". Each task needs what to check, the limit, and what to do when it is out.
- Skipping PMs when production is busy. That is when machines work hardest. Agree the planned stop slot with production in advance.
- Closing PMs without readings. A tick with no reading tells you nothing next month. Record the number, not just "OK".
- Never reviewing intervals. If a PM keeps finding nothing, stretch it. If failures happen between PMs, shorten it or change the task.
- Counting planned PM time as unplanned downtime in OEE. Planned stops come out of planned production time; only unplanned stops reduce availability.
Sources
- Maintenance: Preventive maintenance (opens in a new tab), Wikipedia
- Maintenance of work equipment (opens in a new tab), Health and Safety Executive (HSE), UK
- Total productive maintenance (opens in a new tab), Wikipedia
Frequently asked questions
What is preventive maintenance in simple words?
It is planned work on a machine before it fails, such as cleaning, lubrication, inspection and part replacement, done on a schedule set by time, usage or condition.
What are the main types of preventive maintenance?
Time-based, usage-based, condition-based and predictive. Most plants use a mix, chosen machine by machine according to how each one wears.
How often should preventive maintenance be done?
Start from the OEM manual, then adjust with your own records. If a PM keeps finding nothing, stretch the interval. If failures happen between PMs, shorten it or change the task.
How do you measure preventive maintenance?
Track PM compliance: PMs done within their window divided by PMs due in the period. Read it alongside breakdown hours on the same machines, so you can see whether the PMs are the right ones.
Does every machine need preventive maintenance?
No. A cheap, non-critical item with a spare on the shelf can run to failure by choice. Put your PM effort on machines that stop the line, are hard to repair or have no standby.
What is the difference between preventive and predictive maintenance?
Preventive maintenance is planned on a time or usage interval. Predictive maintenance uses trends in condition readings, such as vibration or temperature, to decide when the work is needed.