MTTR (Mean Time To Repair) measures maintainability: the average time it takes to repair an asset and return it to service. Lower MTTR means faster recovery from failures.
This guide covers the formula, what to count as repair time, the phases hidden inside every repair, a worked example, how MTTR and MTBF set availability, and where the minutes usually go.
The MTTR formula
MTTR = Total Repair Time / Number of Repairs
- Total repair time: the downtime from each failure until the asset is back in production, added up over the period, using one fixed definition.
- Number of repairs: the corrective repairs in the same period. One failure, one repair event.
What MTTR includes: the phases of a repair
Every repair is several waits and jobs in a row. Most teams count all of them, from the moment the asset stops to the moment it runs good parts again:
| Phase | What it covers | Where minutes hide |
|---|---|---|
| Detection and call | The stop is noticed and someone is called | Nobody sees the stop, or the call goes to the wrong person |
| Response | The technician reaches the machine | Technician busy elsewhere, no priority rule |
| Diagnosis | Finding what failed and why | No fault history, unclear alarms, trial and error |
| Waiting for parts | Getting the spare to the machine | Part not stocked, not where the system says, or ordered after the failure |
| Repair | The physical fix | Missing tools or procedure |
| Test and restart | Checks, first good parts, release | Waiting for quality sign-off or a setter |
The R in MTTR is used for repair, restore and recovery. Some teams count only the repair itself (wrench time); others count the whole stop. Both are valid. Write down which phases you count and keep it, or the trend means nothing. The maintenance work order template records both ends of the stop, stopped at and back in production, so the whole stop is on the form.
Step by step
- Define the repair window. Decide what counts in repair time: detection, diagnosis, the fix and the return to service (be consistent).
- Add up repair time. Sum the total time spent repairing the asset over the period.
- Count the repairs. Count the number of repair events in the same period.
- Divide. MTTR = Total Repair Time / Number of Repairs.
A worked example
Over a month, one machine had 5 repairs that took 1.5, 3.0, 0.5, 4.0 and 1.0 hours (example numbers):
| Step | Calculation | Result |
|---|---|---|
| Repair time | 1.5 + 3.0 + 0.5 + 4.0 + 1.0 | 10 h |
| Repairs | count of events | 5 |
| MTTR | 10 / 5 | 2 h |
Look past the average. The middle repair took 1.5 hours; one 4 hour repair, spent mostly waiting for a part, pulls the mean up. Splitting the same 10 hours by phase shows where to act:
| Phase | Hours | Share |
|---|---|---|
| Detection and response | 1.5 | 15% |
| Diagnosis | 3.5 | 35% |
| Waiting for parts | 2.5 | 25% |
| Repair | 2.0 | 20% |
| Test and restart | 0.5 | 5% |
In this example, diagnosis and waiting take 6 of the 10 hours, while the repair itself takes 2. A faster technician would barely move MTTR; faster diagnosis and the right spare on the shelf would.
Our free MTBF and MTTR calculator works out MTTR, MTBF and inherent availability from your operating time, failures and repair time.
MTTR, MTBF and availability
MTBF (mean time between failures) is how long the asset runs between failures; MTTR is how long each failure takes to fix. Together they give inherent availability:
Inherent availability = MTBF / (MTBF + MTTR)
With an MTBF of 98 hours and an MTTR of 2 hours, availability is 98 / 100 = 98.0%. Halve MTTR to 1 hour and it becomes 98 / 99 = 99.0%, the same gain as doubling MTBF to 198 hours (198 / 200). This is availability from failures and repairs only; OEE availability also counts changeovers, waiting and other stops. See how to calculate MTBF and how to calculate availability.
How to reduce MTTR
- Detect and call at once. An automatic alarm or andon call beats a stop someone notices on the next walk past. See what an andon system is.
- Speed up diagnosis. Keep the repair history per machine, clear alarm texts and a troubleshooting guide for each machine type. The troubleshooting directory lists the common causes and checks for over a hundred machine types.
- Stock the critical spares. Decide which parts must be on the shelf before they fail; the spare parts inventory calculator helps size the stock.
- Standardise the common repairs. Written steps and pre-kitted parts for the failures that repeat.
- Train for the top failure modes. Know who can fix what on each shift.
- Measure each phase. Log when the call, arrival, diagnosis, part and restart happen, so you know which phase to attack.
Common mistakes
- Only timing the wrench-time and ignoring waiting for parts or technicians.
- Counting planned maintenance as a repair.
- Changing what is included in repair time between periods, so the trend is meaningless.
- Reporting MTTR without the number of repairs behind it; with a handful of repairs, one long job swings the figure.
- Looking only at the mean, so the few very long repairs that cost the most go unnoticed.
- Mixing assets with very different repair profiles into one MTTR.
MTTR is only as good as the stop times behind it. Fabrico records the start and length of every stop live from the machine, so restore time per stop and per machine comes from data, not memory.
MTTR FAQ
Is a lower MTTR better?
Yes. A lower MTTR means faster repair and return to service, so less downtime per failure.
What should MTTR include?
Be consistent. Many teams include detection, diagnosis, the physical repair and the return to service. Waiting for spares is a big, often hidden, part.
How does MTTR affect availability?
Inherent availability is MTBF / (MTBF + MTTR), so every cut in MTTR raises availability, because each failure costs less downtime.
What is the difference between MTTR and MTBF?
MTBF is the average running time between failures and measures reliability. MTTR is the average time to repair a failure and measures maintainability. You need both to understand availability.
Does MTTR include planned maintenance?
No. MTTR covers repairs after failures. Planned maintenance is tracked separately, for example as preventive maintenance compliance.
What is a good MTTR?
There is no universal number: it depends on the asset, the definition you use and how many repairs it is based on. Compare each asset with its own trend and always report the number of repairs next to MTTR.
Related: MTBF / MTTR calculator · MTBF · planned vs unplanned downtime · cost of downtime