MTBF and MTTR explained for factory managers
Ask a factory manager how reliable the machines are and you will usually hear an impression: "the overlocks are always breaking" or "line three is fine". Impressions are useful, but they cannot tell you whether things are getting better or worse, or whether the problem is the machines or the maintenance team. Two simple numbers can: MTBF and MTTR.
Both come from maintenance engineering, but neither needs an engineer to calculate. If you know when each breakdown started and when the machine was back in service, you have everything you need.
MTBF: how often machines break
MTBF stands for mean time between failures. It is the total running time of a machine, or a group of machines, divided by the number of breakdowns in that period.
Example: a flatlock machine works two shifts of 8 hours, 25 days a month, for 400 scheduled hours. Subtract 10 hours of downtime and it ran 390 hours. It broke down 5 times. Its MTBF is 390 ÷ 5 = 78 hours, meaning on average it runs for about five shifts between failures.
Higher is better. A falling MTBF tells you something is changing in the machine itself or in how it is used: wear, poor lubrication, a new fabric that is too heavy for the setup, or an operator forcing the machine.
MTTR: how long machines stay broken
MTTR stands for mean time to repair. It is total downtime divided by the number of breakdowns. Using the same flatlock: 10 hours of downtime over 5 breakdowns gives an MTTR of 2 hours.
Lower is better. Note that MTTR measures the full time the machine was unavailable, from the moment it stopped to the moment it was back in production, not just the minutes the mechanic spent with a screwdriver. That distinction matters, because on most floors the repair itself is quick and the waiting is not.
A rising MTTR almost never points to the machine. It points to the response:
- Breakdowns are reported late because the operator does not know whom to call.
- Mechanics are spread too thin across too many lines.
- Common spare parts such as needles, loopers, knives and belts are not in stock.
- The fault is misdiagnosed, so the machine is "fixed" and stops again an hour later.
Reading the two together
Each metric alone can mislead. Together they tell you where to act:
| MTBF | MTTR | What it usually means |
|---|---|---|
| High | Low | Healthy machine, good response. Leave it alone. |
| Low | Low | Frequent small stoppages, such as thread breaks and skipped stitches. Check settings, needles and operator habits. |
| High | High | Rare but serious failures. Check spare parts stock and mechanic skills for that machine type. |
| Low | High | The machine is wearing out and repairs are slow. Consider replacement. |
The last row is where the lifecycle question begins. If a machine has sat in that corner for several months, read the machine lifecycle from purchase to replacement before spending more on repairs.
Turning the numbers into availability
If you want one figure for the owner, combine the two into availability: MTBF ÷ (MTBF + MTTR). For the flatlock above, 78 ÷ (78 + 2) = 97.5 percent. That sounds good until you notice that 2.5 percent of 400 hours is 10 hours, more than a full shift lost every month on one machine. Multiply that across a line of 30 machines and the cost becomes very visible.
Availability also explains why a line's real output rarely matches its theoretical capacity. When you calculate line efficiency, as described in how to calculate SMV and line efficiency, machine downtime is one of the first gaps to look at.
Getting the data right
MTBF and MTTR are only as honest as the timestamps behind them. A few practical rules:
- Record the start time when the machine stops, not when the mechanic arrives. Otherwise you hide the waiting time, which is often the largest part.
- Record the end time when production resumes. A machine that is repaired but waiting for a test seam is still down.
- Record at the machine. Breakdowns written up at the end of the day are estimated, and estimates are always rounder and shorter than reality.
- Choose a reason from a fixed list. This lets you see which causes drive the numbers, as explained in Pareto analysis of machine breakdown causes.
- Review monthly, not daily. One bad day moves the numbers sharply; trends over months are what matter.
Where Nasij fits
In Nasij, every malfunction is logged against a specific machine with a start time, an end time, its line, and one or more predefined reasons. Supervisors open the machine record by scanning its QR sticker with any phone, so the start time is captured at the machine when the stoppage happens, not reconstructed later. With that history in one place, MTBF and MTTR stop being abstract and become two figures you can work out each month from the log and compare between machines, lines and machine types. They also feed directly into decisions about which machines need preventive rather than corrective maintenance.
