MTTR: What Is It? Formula, Example, and Ways to Reduce Downtime
Every minute of downtime costs money, but the failure itself is not always the biggest problem. The bigger issue is the lack of control over repair time.
MTTR, or Mean Time To Repair, is a metric that shows the average time needed to repair a machine, device, or system after a failure. In maintenance, it helps show how long it takes, on average, to bring equipment back to working condition after a stop.
For a maintenance manager, production leader, or plant director, MTTR gives useful information: whether the team responds fast, whether it has access to spare parts, repair history, and the data needed to make decisions.
IBM defines MTTR as the average repair time after a failure, calculated by dividing total repair time by the number of repairs completed in a given period.
MTTR: What Is It?
MTTR stands for Mean Time To Repair, or average repair time. In manufacturing, the MTTR metric shows how long it usually takes to handle a failure from the start of the repair until the machine returns to work.
You can calculate MTTR for one machine, a full line, a group of devices, a specific failure type, a shift, or an entire plant. The most useful view is usually at the level of specific machines and faults. A company-wide average can hide the problem. If one line returns to work after 30 minutes and another after 4 hours, one shared result does not explain much.
MTTR Mean Time To Repair: What Exactly Should You Measure?
The MTTR metric needs a clear definition first. The abbreviation is used in several ways. Atlassian notes that MTTR may stand for Mean Time To Repair, Mean Time To Recovery, Mean Time To Restore, Mean Time To Resolve, or Mean Time To Respond.
In maintenance, MTTR most often means Mean Time To Repair, or the repair time itself. In a broader view, companies measure the total unavailable time, from the failure to the moment the machine returns to normal work.
The rule is simple: choose one definition and use it every time. Otherwise, MTTR may look good in a report but will not help in a discussion with production.
It is useful to measure separately:
- time from failure to report,
- time from report to work order acceptance,
- diagnosis time,
- repair time,
- testing and restart time,
- full production downtime.
You can then see whether the problem is the repair itself, a late report, missing spare parts, or a long diagnosis.
MTTR Metric: Formula
The basic formula is simple:
MTTR = total repair time / number of repairs
If a machine had 10 repairs in a given month and they took 500 minutes in total, the MTTR is 50 minutes.
This calculation matches IBM’s definition: you divide the total time spent on repairs by the number of repairs completed in the period you are analyzing.
MTTR Calculation Example
Assume that in May a given machine had 5 failures. The repairs took: 45 minutes, 90 minutes, 30 minutes, 120 minutes, and 75 minutes.
Total repair time was:
45 + 90 + 30 + 120 + 75 = 360 minutes
Now divide the result by the number of repairs:
MTTR = 360 minutes / 5 repairs = 72 minutes
This means the average repair of this machine took 72 minutes.
The number alone is not enough. You still need to check whether 72 minutes is an acceptable result, how much one hour of downtime for this machine costs, and where time is lost: reporting, diagnosis, parts, repair, or restart after repair.
How Much Does High Mean Time to Repair Cost?
The cost of downtime depends on the industry, margin, production load, material losses, labor, and delivery dates. The scale can vary widely.
The Siemens report “The True Cost of Downtime 2024” states that one hour of unplanned downtime in a large automotive plant can cost $2.3 million, or more than $600 per second.
Not every plant operates on that scale. The mechanism is still the same: the longer the repair takes, the longer the line does not produce.
Example for a mid-sized manufacturing company:
- cost of one hour of line downtime: PLN 4,000,
- number of failures per month: 20,
- current MTTR: 90 minutes,
- target MTTR: 60 minutes.
Reducing MTTR by 30 minutes across 20 failures gives 10 fewer hours of downtime per month.
At PLN 4,000 per hour, that means:
10 hours × PLN 4,000 = PLN 40,000 per month
Over a year, that gives PLN 480,000 in potentially recovered production time. This is a sample calculation, not a promised result. It shows why MTTR should be linked with downtime cost.
How to Interpret Mean Time to Repair
Low MTTR usually means the team restores equipment quickly after a failure. High MTTR may point to a problem in the process, data, or work organization.
The most common causes of high MTTR are:
- failures reported too late,
- incomplete problem descriptions,
- no history of previous repairs,
- missing spare parts,
- long diagnosis,
- no instructions for recurring faults,
- production and maintenance using different downtime calculations,
- treating symptoms instead of root causes.
MTTR is best compared over time and broken down by machines, failure types, lines, and shifts. One weaker month may come from one major failure. A steady increase over several months means you should check how repairs are reported and handled.
MTTR vs. MTBF: What Is the Difference?
Mean Time to Repair shows how long a repair takes after a failure. The lower the MTTR, the faster the machine returns to work.
MTBF, or Mean Time Between Failures, shows the average operating time between failures. The higher the MTBF, the less often the machine fails.
IBM describes MTBF as the average time a system or product operates before failure, and MTTR as the average time needed to repair it after failure. MTBF does not include repair time, and MTTR does not show the time between failures.
Example: machine A fails rarely, but each repair takes 4 hours. Machine B fails more often, but returns to work after 25 minutes. Only a combined look at MTTR and MTBF shows whether the bigger problem is failure frequency or repair time.
Let’s talk about how to reduce MTTR at your plant.
Want to check how you can reduce MTTR in your production process, but you are not fully sure how to prepare? Contact us, and we will help you assess the current state and plan the next steps.
How to Reduce MTTR
The biggest gains come from actions that shorten the path from failure to machine restart.
Start with these areas:
- Organize failure reports. A report should include the machine, location, description, priority, photo, and time when the problem occurred.
- Separate response time from repair time. You will see whether the team spends too long repairing, or too long waiting for information, parts, or a decision.
- Keep repair history linked to each machine. Previous failures, replaced parts, and technician comments shorten diagnosis.
- Control spare parts availability. A missing part can turn a short repair into hours of downtime.
- Analyze recurring failures. If the same fault returns every few days, a fast repair does not solve the problem. You need to find the cause.
- Measure MTTR from system data, not notes. Paper notes, calls, and spreadsheets make analysis harder. Data should be collected in one place.
How Does CMMS Help Control MTTR?
If you calculate MTTR in a spreadsheet today, or recreate the course of a failure from calls and notes, you first lose time collecting data and only then move to analysis. A CMMS helps organize the process: repair requests, statuses, repair history, parts, and reporting are kept in one place.
A well-set CMMS gives you:
- one place to report failures,
- status of every report,
- start and end time of each repair,
- machine history,
- spare parts information,
- data on recurring faults,
- MTTR and MTBF reports based on real events.
A CMMS collects service data in one place and helps you make decisions based on real metrics, such as MTTR and MTBF.
It is a natural next step for companies that want to respond faster to failures and better understand where downtime starts.
Common Mistakes When Calculating MTTR
- The most common mistake is calculating one average for the whole plant. It is better to break results down by machines, lines, failure types, and shifts.
- The second problem is mixing repair time with full downtime. If production waits for an operator, a quality test, or a changeover, you need to clearly define whether that time is included in MTTR.
- The third mistake is the lack of one shared definition. If one shift counts time from the report and another from the start of repair, the result cannot be compared.
What to Do Next
If you are responsible for maintenance or production, start with one line or one group of machines. For 30 days, measure MTTR using one definition.
Check:
- which machines have the highest MTTR,
- which failures return most often,
- where the most time is lost,
- how much one hour of downtime costs,
- whether technicians have access to repair history and spare parts.
After 30 days, you will see whether the biggest problem is reporting, diagnosis, parts, repair, or restart. That is enough to stop talking about failures in general terms and start improving a specific part of the process.
If data is spread across many places, the next step should be moving reports, repair history, and MTTR reporting into one CMMS.
FAQ: What Is MTTR?
What does MTTR mean?
MTTR means Mean Time To Repair, or the average repair time after a failure.
How do you calculate MTTR?
You calculate MTTR with this formula: total repair time / number of repairs.
Is low MTTR always good?
Usually yes, but not always. If a machine fails very often, low MTTR may only hide the problem. In that case, you should also check MTBF.
What is the difference between MTTR and MTBF?
MTTR tells you how long a repair takes. MTBF tells you how long a machine works between failures.
What is a good MTTR?
There is no single value for every company. A good MTTR depends on machine type, failure complexity, downtime cost, and spare parts availability. It is best to compare the result month to month and between similar machines.
Let’s reduce MTTR in your production.
See how else you can improve production with articles on the explitia blog.
