Every maintenance strategy is a mix of two things: work you plan before a failure, and work you do because something already failed. The question is never "preventive or corrective?" — it is "what is the right mix for each asset?"
This article defines the main maintenance strategies, compares their costs and risks honestly, walks through a worked cost comparison, and shows how to decide which assets deserve a preventive plan.
Definitions
Corrective maintenance is work done to restore an asset after a fault has been found or it has failed. It includes emergency breakdown repairs, but also planned corrections of defects found during an inspection.
Preventive maintenance (PM) is work done on a schedule, before failure, to reduce the likelihood of failure. The schedule can be time-based (every 90 days), usage-based (every 500 operating hours) or a combination. See the glossary entry on preventive maintenance.
Two related strategies are worth knowing:
- Condition-based maintenance performs work when a measured condition — vibration, temperature, oil analysis, wear — crosses a threshold.
- Predictive maintenance uses trends in condition data to estimate when a failure is likely and schedule work just before it.
And one deliberate strategy that is often overlooked:
- Run-to-failure is a conscious decision to do no preventive work on an asset because failure is cheap, safe and quick to fix. A desk lamp is the classic example.
The honest comparison
| Corrective (reactive) | Preventive | |
|---|---|---|
| When work happens | After a fault or failure | On a schedule, before failure |
| Planning | Little; urgent and disruptive | Planned, with parts and people ready |
| Downtime | Unplanned, often long | Planned, usually shorter |
| Labour cost | Often higher (overtime, call-outs) | Predictable |
| Parts cost | Expedited shipping, secondary damage | Standard ordering |
| Risk of over-maintenance | None | Real — some tasks may be unnecessary |
| Safety and compliance | Higher risk | Demonstrable care, records for inspectors |
Preventive maintenance is not free. Every PM task costs labour, parts and some downtime, and a poorly designed PM program can replace components that had plenty of life left. The goal is to do the tasks that genuinely prevent expensive failures, and no others.
A worked cost comparison
Illustrative example with assumed numbers. Replace them with your own.
An air compressor supports a small production line.
Run-to-failure scenario (per year):
- Expected breakdowns: 2
- Repair cost per breakdown (parts, call-out, labour): $1,800
- Lost production during each breakdown: 6 hours × $400 per hour = $2,400
- Annual cost: 2 × ($1,800 + $2,400) = 2 × $4,200 = $8,400
Preventive scenario (per year):
- Quarterly service: 4 × $350 = $1,400
- Planned downtime per service: 1 hour, scheduled outside production, so $0 lost output
- Expected breakdowns reduced to 0.5 per year: 0.5 × $4,200 = $2,100
- Annual cost: $1,400 + $2,100 = $3,500
In this example, PM saves $8,400 − $3,500 = $4,900 per year. The key sensitivities are the cost of lost production and how much the PM tasks actually reduce the failure rate. If the compressor had a standby unit, lost production would be close to zero, and run-to-failure would cost 2 × $1,800 = $3,600 — almost the same as the preventive plan. That is why the decision has to be made asset by asset.
Which assets deserve a PM plan?
Score each asset or asset class on four questions:
- Consequence of failure. Does it stop production, create a safety hazard or breach a regulation?
- Failure pattern. Does it wear out predictably (belts, filters, bearings)? PM works best on wear-related failures. Random electronic failures are rarely prevented by time-based servicing.
- Cost of the PM task relative to the failure it prevents.
- Redundancy. Is there a standby unit or an easy workaround?
High consequence plus predictable wear equals a strong PM candidate. Low consequence plus random failure equals run-to-failure. Everything in between is a judgement call — start with manufacturer recommendations and adjust based on your own failure history.
Building a basic PM schedule
For each PM candidate, define:
- The task — what exactly is done (inspect, lubricate, replace filter, calibrate).
- The trigger — calendar interval, usage interval, or both ("every 3 months or 500 hours, whichever comes first").
- The skill and time required.
- Parts and consumables.
- The acceptance criteria — what "pass" looks like, and what to do if it fails.
Our guide to building a preventive maintenance program covers scheduling, backlog management and metrics in depth.
Measuring the balance
A handful of metrics tell you whether your mix is right:
- Planned vs unplanned work ratio — the share of maintenance hours spent on scheduled work.
- PM compliance — the share of PM tasks completed on time.
- Mean time between failures (MTBF) for critical assets — should rise as PM takes effect.
- Maintenance cost per asset — to spot assets that cost more to keep than to replace.
None of these metrics is possible without a service history attached to each asset. That is the first thing to fix if you are purely reactive today: log every repair as a work order against the asset.
Maintenance in Asetavo
Asetavo's maintenance module links every work order — with status, cost and parts — to the asset, so the Maintenance tab shows its complete history. Preventive schedules produce a due list of upcoming work today; automatic work-order generation from schedules is on the roadmap. Calibration is logged through work-order completion, and reports show maintenance cost by asset so repair-or-replace decisions use real data. Maintenance is part of the Business plan — see pricing.
Related reading: CMMS in the glossary, and what is asset management?