The two indicators are always quoted together and measure opposite things: how often it breaks, and how long it takes to get back to producing. They improve through different actions, and there is one recording mistake that invalidates both. With a calculator to decide whether the preventive plan pays for itself.
MTBF and MTTR are quoted together and measure opposite things. The first says how often it breaks; the second, how long it takes to get back to producing. One is about reliability, the other about response capability, and they improve through different actions.
MTBF and MTTR count BREAKDOWN stoppages only. Planned maintenance is not a failure: if its hours go into MTTR, the indicator stops measuring reliability and starts measuring how much maintenance you do — the exact opposite of what you want to know. Preventive hours are counted separately, and they do weigh in the economic balance, as shown below.
The question that settles a maintenance budget is not "what does the plan cost?" but "what does not having it cost?". The starting values are those of an ordinary production line; replace them with yours.
—
All three coexist: this is not about replacing one with another but about deciding which equipment belongs in each regime. Putting everything on predictive is as expensive as leaving everything on corrective.
Once it has already failed. It costs nothing until it costs a lot. It is the right choice for equipment whose failure does not stop the line and whose repair is cheap and quick.
Every so often, by time or by usage, whether it fails or not. You pay for interventions that were sometimes unnecessary, in exchange for the failure not picking the moment. It is the default regime for anything critical.
When the measurement says failure is approaching: vibration, temperature, power draw. It needs instrumentation and somebody to read the data. It only pays for itself on equipment that is very critical or very expensive to stop.
A machine moves up a regime when the expected cost of its failure exceeds the cost of watching it. That comparison is the one in the calculator above, done machine by machine — not in one go for the whole plant.
ABC criticality ranks the asset base by the consequence of failure, not by the price of the machine. A €1,200 pump that stops the whole line is more critical than a €40,000 compressor with a twin beside it.
| Class | What happens if it fails | Sensible regime | Spares |
|---|---|---|---|
| A · critical | Stops production, affects safety or breaks a customer commitment. No alternative. | Preventive on a firm schedule; predictive if the cost of downtime justifies it. | On site, with a defined minimum stock. |
| B · important | Reduces capacity or quality, but there is a way to carry on: an alternative route, a slower rate, recoverable. | Preventive with slack in the dates; inspection during planned stops. | Long-lead items on site; the rest to order. |
| C · secondary | It is a nuisance and it gets fixed. It compromises neither schedule nor quality. | Corrective. A preventive plan here is spending with no return. | To order. |
The classification is reviewed when the line changes, not when the machine changes: a machine goes from C to A the day the alternative that covered it disappears.
The calculation above is simple arithmetic on data that has to be recorded properly. What has real depth is how each machine’s regime is decided.
The European standard that fixes the vocabulary: corrective, preventive, condition-based, predetermined. Useful because a good share of maintenance arguments are arguments about what each word means.
Defines a set of economic, technical and organisational indicators. Its main use is to stop you inventing a definition for a KPI that already has one.
The RCM reference. Its central contribution: the maintenance plan is derived from the equipment’s functions and failure modes, not from the manufacturer’s catalogue.
FMEA applied to production equipment. It is the bridge between the recorded breakdown log and the maintenance plan, because it forces you to ask about the cause before setting a frequency.
The origin of OEE and autonomous maintenance. Its contribution is the idea that the person operating the machine is the first to notice something is wrong, long before any indicator does.
No method fixes an incomplete log. If the stop time, the restart time and the cause are not recorded, MTBF and MTTR are estimates, and it is worth saying so when you present them.