Blog / Reliability
Reliability · Guide

Mean Time Between Stops: Formula and Example

OL By OEE Lab Editorial |Updated 13 November 2026 |6 min read
Short answer

Mean time between stops (MTBS) is the average run time between one stop and the next. The formula is MTBS = runtime / number of stops. A higher MTBS means the line stops less often. It measures frequency, while mean stop duration measures how long each stop lasts.

Mean time between stops calculator

Enter one shift or period. The defaults are the worked example below. Updates as you type.

Run time
min
Shift minus planned breaks.
min
All the stops you count, added up.
Stops
Every stop longer than your counting rule, for example 10 seconds.
Optional for the share of short stops
How many of the stops were short.
s
Stops under this length count as short.
Defaults: 480 min, 24 stops
Mean time between stops
- min
MTBS = run time ÷ number of stops
0%100% availability
Run time
-
Stops
-
Stops per run hour
-
Mean stop length
-
Availability
-
Short stops
-

The average run and stop

Running-
Stopped-

The working

MTBS counts every stop, MTBF counts only equipment failures. Work it out with the MTBF and MTTR calculator.

What MTBS tells you

Availability tells you how much time you lost. MTBS tells you how that loss is made up.

Ten stops of 6 minutes and 60 stops of 1 minute both cost an hour. They are not the same problem.

The first needs better repairs. The second needs a cause fix at the machine.

MTBS is the frequency half of the story. The duration half is the mean stop duration.

The formulas

MTBS = runtime / number of stops
Mean stop duration = total stop time / number of stops
Availability = MTBS / (MTBS + mean stop duration)
MTBS, mean stop duration and availability

The third line is the same as runtime / scheduled time, written from the two averages. It works when every stop inside scheduled time is counted.

Use runtime in the first line, not scheduled time. The machine cannot stop while it is already stopped.

Worked example

Example: a line has 480 scheduled minutes. It stops 24 times, for 72 minutes in total.

StepResult
Runtime480 - 72 = 408 min
MTBS408 / 24 = 17.0 min
Mean stop duration72 / 24 = 3.0 min
Availability17 / (17 + 3) = 85.0%
Check: runtime / scheduled408 / 480 = 85.0%

Both routes agree. The line runs on average 17 minutes, then stops for 3.

Two ways to the same gain

Now halve the loss. You can halve how often it stops, or how long each stop lasts.

Availability moves the same way. MTBS does not.

CaseStopsMean stopMTBSAvailability
Today243.0 min17.0 min85.0%
Half the stops123.0 min37.0 min92.5%
Half the duration241.5 min18.5 min92.5%

Check "half the stops": runtime is 480 - 36 = 444 minutes, and 444 / 12 = 37.0. Check "half the duration": runtime is 444 minutes again, and 444 / 24 = 18.5.

Today17.0 min
Half the stops37.0 min
Half the duration18.5 min
MTBS in the three cases. Today 17.0 minutes, half the stops 37.0 minutes, half the duration 18.5 minutes.

Same availability gain, very different MTBS. So MTBS tells you whether your fix reduced the number of stops. Availability alone cannot.

Which one is easier? If the stops are short and frequent, cut the number. If they are rare and long, cut the duration, with spares, staging and a better response.

What it does to the number when you count small stops

MTBS depends on the smallest stop you count. This is the biggest trap.

Example: the same shift, but a counter also finds 40 more stops of 30 seconds (20 minutes) that nobody logged.

ViewStopsStop timeMTBS
Logged stops only2472 min17.0 min
Plus 30-second stops6492 min6.1 min

Runtime falls to 388 minutes, and 388 / 64 = 6.1. Availability is 388 / 480 = 80.8%.

The line looked like it ran 17 minutes between stops. It really ran about 6. These are the micro-stops that never reach the log.

So fix a rule: count every stop longer than a set length, for example 10 seconds, and say so on the chart. Then the trend is comparable from week to week. The rule is easy to hold when stops are read from the machine, and hard when they are written by hand.

Reading the weekly trend

Put MTBS and mean stop duration next to each other, one row a week. Example:

WeekMTBSMean stopAvailability
117.0 min3.0 min85.0%
218.0 min3.0 min85.7%
312.0 min3.0 min80.0%
412.0 min4.0 min75.0%

In week 3 the line stopped more often, with the same length of stop. In week 4 the stops also got longer.

A single availability number would show a slide. The two columns show what slid, and which team has to act: the one that prevents stops, or the one that restores the line.

MTBS, MTBF and MTBI

These three are close relatives. They differ in what they count.

The repair side has two terms in the same book. MTTR is repair time, from the mechanic's arrival.

MTTRs is restore time: from the end of production to the start of the next, including the call, travel, diagnosis, parts and testing. Define the one you mean before you compare. See how to calculate MTTR.

The same book shows a case where output fell about 10% with no big event behind it. The average time to restore the line after each tiny stop had crept from 0.8 to 1.1 minutes, and only a chart of operational MTTR (mean restore time per stop) showed it.

So plot both MTBS and mean stop duration, every week.

How to raise MTBS

  1. 1Count every stop. Read stops from the machine if you can. Hand logs miss the short ones.
  2. 2Chart the number of stops by cause. A Pareto by count, not only by minutes. See the downtime Pareto chart.
  3. 3Attack the top cause at the source. Jams, sensors, feeding, settings. Ask why until you reach something you can change.
  4. 4Check MTBS after each fix. It should rise. If not, the stops moved to another cause.
  5. 5Hold the gain. Update the standard, train the crews and watch the trend.
Five steps to fewer stops

The causes of frequent short stops are listed in how to reduce micro-stops.

See it live in Fabrico

MTBS needs a complete stop count, including the short ones. Fabrico takes stops, speed and counts live from PLCs or IoT sensors (external sensors on older machines), with add-on AI cameras counting output where a machine has no usable signal. See how Fabrico counts every stop.

Book a Fabrico demo

What people ask next

What is a good MTBS?

There is no number that fits every line. It depends on the process, the product and what you count as a stop. Compare each line with its own past and watch the trend.

Should planned stops count in MTBS?

It is your choice, but state it. If you count only unplanned stops, you are close to the MTBI in Overall Equipment Effectiveness (R. C. Hansen, 2001). If you count all stops, your MTBS also shows changeover frequency.

Is MTBS the same as MTBF?

No. MTBF counts equipment failures only. MTBS counts every stop you decide to count, whatever the cause. See how to calculate MTBF.

How is MTBS linked to availability?

Availability = MTBS / (MTBS + mean stop duration). A higher MTBS or a shorter stop both lift availability. See how to calculate availability.

Do I divide by scheduled time or runtime?

Use runtime, because it is the time the line was actually running between stops. Whichever you choose, keep it the same every week.

Related guides: How to Calculate Uptime: Formula and Examples · MTBF for Multiple Machines: Series and Parallel · What Is a Good MTBF? How to Set Your Own Target

← More OEE Lab guides · Open the OEE Calculator →

OEE essentials: OEE calculator · What is OEE · Six big losses · Availability · TEEP · Cost of downtime · OEE benchmark · Free OEE tracker · Free OEE software · How to track downtime · How to implement OEE · Light MES vs full MES