Blog / Downtime
Downtime · Guide

How to Set a Downtime Target for a Production Line

OL By OEE Lab Editorial |Updated 2 November 2026 |6 min read
Short answer

Set a downtime target from your own line's baseline, not from an outside figure. Take three to six months of stops, split them by cause, and estimate how many minutes each planned fix removes. Your target is the baseline minus those minutes, for example from 96 to 70 minutes per shift by the end of the quarter, with an owner per cause.

"Reduce downtime by 20%" sounds like a target, but it is really a wish. It does not say which stops, by when, or who has to do what.

A target that works reads like a budget. It lists each cause, the minutes it may use, and the action that gets it there.

If your goal is an OEE number, start with how to set OEE targets. This page covers the stop minutes for a single line.

Step 1: get a baseline you trust

Three to six months of stop data is a good basis, and less can work if you label the target as provisional. Check the data first, using how to check downtime data.

A target set on bad data has two sad endings. Either you hit it by logging less, or you miss it because the old log was flattering you.

Write down the rules with the baseline: what counts as a stop, your stop threshold, and what is excluded from scheduled time. The target is valid only under those rules.

Step 2: split it by cause

One total number gives nobody anything to do. Separate planned from unplanned stops, then rank the causes with a downtime Pareto chart.

Example: a line with 450 scheduled minutes per shift, over three months. Stops average 96 minutes per shift, so the downtime percentage is 96 / 450 = 21.3%.

Step 3: count what each fix removes

This step is what makes the target honest. For each big cause, name a real action and estimate the minutes it removes, then let the person doing the work challenge that estimate.

CauseTodayActionTargetOwner
Changeovers (planned)34 minStage tools before the stop, use a settings sheet26 minLine lead
Breakdowns22 minFix the top two failure modes16 minMaintenance lead
Waiting for material18 minStaging at shift start, one feed route10 minLogistics
Logged jams14 minRework the infeed guide10 minProcess engineer
Other8 minNone this quarter8 min
Total96 min70 min
The target as a budget, in minutes per shift. Every number and action here is an example.
Today96 min
Changeovers-8 min
Breakdowns-6 min
Waiting for material-8 min
Logged jams-4 min
Target70 min
From 96 to 70 minutes per shift. Full bar = 96 minutes.

Notice "other" gets no reduction. A cause with no action does not earn a target, because a number with no plan behind it will just be missed.

If "other" or "unknown" is large, fixing the coding is the first action. See unknown downtime.

Step 4: check it against OEE and output

A stop target means little until you see what it does to the line. So hold performance and quality steady, and run the formula.

Today: (450 - 96) / 450 = 78.7% availability
78.7% x 90% x 98% = 69.4% OEE
Target: (450 - 70) / 450 = 84.4% availability
84.4% x 90% x 98% = 74.5% OEE
With performance at 90% and quality at 98% in the example, the stop target is worth 5.1 OEE points.

If your OEE target needs more than that, this plan is not enough, and the rest must come from speed or quality.

Then ask what those 26 minutes are worth. The downtime cost calculator gives you that number in money, which helps decide how much a fix may cost.

Step 5: start where it counts

Downtime on the constraint costs output directly. On a machine with spare capacity it may cost much less, since the time can be made up.

Set tighter targets on the constraint first. Bottleneck analysis shows how to find it.

Should you use an outside benchmark?

Use one for context, and not as your target. One neutral figure comes from a NIST survey of US discrete manufacturers (2016 data), where unplanned downtime took 7.8% of planned production time.

In the same report, the most reactive quarter of plants lost 13.0%, against 4.0% for the least reactive. See manufacturing downtime statistics for more.

That figure is an average across many industries, with survey answers and its own definitions. A line on a high-mix schedule and a dedicated continuous line are not comparable, so your own baseline stays the better guide.

If you do compare, compare like with like. Our example's unplanned part is 62 of the 96 minutes, or 13.8%, once planned changeovers are taken out.

Write it so it can be managed

  • Use minutes per shift. People can picture 16 minutes. A percent is harder to act on.
  • Set a date. For example, the end of the quarter, with a midpoint check: 96 to 85 minutes by week 6, and 70 by week 13.
  • Name an owner per cause. A total owned by everyone is owned by nobody.
  • Review weekly. Look at the trend by cause, not at a single bad day.
  • Reset when you get there. Once a target holds for a month or two, redo the Pareto and set the next one.

How targets go wrong

Any stop target can be gamed, usually without bad intent. Put these checks beside it from the first day.

What can happenThe check that catches it
Stops get logged less, or get shorter on paperCompare with machine run time and counts
Stops are cleared just under the thresholdTrack stop count and performance next to minutes
Planned maintenance gets skipped to save minutesTrack maintenance done against plan, and failures in later months
Changeovers get rushed, and scrap goes upTrack the changeover to first good unit, and scrap in that window
Long runs replace changeovers, and stock goes upTrack stock and changeover frequency
Pair every downtime target with OEE and the stop count, so a win in one number cannot hide a loss in another.

What if the line has no data?

Do not invent a number. Log every stop for four weeks with a downtime log template, set a provisional target, and redo it after three months.

See it live in Fabrico

A budget by cause needs stop minutes you can trust. Fabrico takes stops live from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, operators tag the cause at the line, and you see OEE per line and shift. See how Fabrico tracks stops by cause.

What people ask next

What is a good downtime percentage?

It depends on the process, the mix and what you count, so a single good number would mislead you. A NIST survey of US discrete manufacturers (2016 data) found unplanned downtime took 7.8% of planned production time, on average. Learn the math in how to calculate downtime percentage.

Should planned downtime have a target too?

Yes, but a different kind. Set planned stops by standards, such as minutes per changeover, and do not push them to zero. See planned vs unplanned downtime.

Downtime target in minutes or in percent?

Use minutes per shift on the floor, because people can act on them. Report the percent upward if your plant prefers it. The two convert easily using scheduled time.

How much can we cut downtime in a year?

We found no neutral figure that applies to any plant, so build it bottom-up from named actions. Set a target for one quarter, see how well your estimates held, then plan the next. How to reduce unplanned downtime has the actions.

Who should own a downtime target?

Give the line lead the total, and give each cause to the person who can act on it. Maintenance cannot own waiting for material, and logistics cannot own bearing failures. Review them together at the daily production meeting.

See OEE per line and shift on your own machines

Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.

Book a Fabrico demo

Related guides: OEE Improvement Action Plan: A 90-Day Template · Unplanned Downtime Formula and KPI Calculation · What Is a Good MTBF? How to Set Your Own Target

← More OEE Lab guides · How to set OEE targets →

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