Blog / Downtime
Downtime · Guide

Machine Downtime Reasons: 12 Causes to Check

OL By OEE Lab Editorial |Updated 2 November 2026 |5 min read
Short answer

The usual machine downtime reasons are equipment failures, jams, sensor faults, changeover overruns, tool changes, no material, no operator, blocked or starved lines, quality holds, cleaning overruns, startup problems and utility cuts. We found no neutral ranking that holds across plants, so your own stop data has to set the order.

What published studies can and cannot tell you

Any "top reason" list without a source is a guess. Two neutral sources do say something useful.

In a NIST survey of US discrete manufacturers (2016 data), 31.7% of unplanned downtime was due to maintenance issues. So in that survey, most unplanned downtime was put down to something else.

In a study of 884 machines in 23 Swedish companies (Hedman et al., 2016), unclassified losses were about 19% of scheduled time. That was more than half of all recorded loss time, so the reason was missing or too vague to use.

Do not assume the machine is the problem, and expect a big "nobody knows" bar until you fix it. See unknown downtime.

Twelve reasons and what to check first

ReasonCheck firstGroup
Mechanical failureIs it the same part as last time?Machine
Electrical or drive faultThe fault code on the drive or panelMachine
Sensor fault or false tripDirt, alignment, a loose cableMachine
JamGuide settings, product and pack variationMachine
Tool change or wearPlanned change or early failure?Process
Changeover overrunWhich step ran past its standard?Process
Startup problemsSettings and temperature at restartProcess
Quality holdWhat triggered the hold, and who releases it?Process
No materialWas it late, or in the wrong place?Supply
No operatorBreaks, absence, shared crewSupply
Blocked or starvedWhich machine is the real constraint?Supply
Utilities or cleaning overrunAir, power, water, or a clean that ran longSite
The first check is a starting point for the person at the line, not a diagnosis.

Several of these have their own guide: sensor fault stops, tool change downtime, waiting for an operator and startup losses. For the fault itself, the troubleshooting directory lists symptoms by machine.

Only some of these reasons are breakdowns. If a list of reasons is all machine faults, the supply and process stops are hiding somewhere.

Minutes and count give two different top reasons

Example: one week on a packaging line with 2,400 planned minutes. The operators and a stop counter record these reasons.

Mechanical failure (5 stops)210 min
Jams (62 stops)124 min
No material (8 stops)96 min
Changeover overrun (6 stops)72 min
Quality holds (3 stops)63 min
Waiting for operator (9 stops)54 min
Minutes by reason, with stop counts in the labels. Failures lead on minutes from just 5 stops.

The total is 619 minutes across 93 stops. Availability is (2,400 - 619) / 2,400 = 74.2%.

Mechanical failure leads on minutes, at 210 of 619, or 33.9%. Jams lead on count, at 62 of 93 stops, or 66.7%.

Each ranking points to a different job. Failures ask for a repair plan and spares, and jams ask for a fix at the machine.

Put each reason in OEE points

Minutes turn into a decision once you divide by planned time. On this line, each 24 minutes a week is one point of availability.

  • Cut jams by half: 62 minutes back, or 62 / 2,400 = 2.6 points of availability.
  • Cut failure minutes by half: 105 minutes back, or 4.4 points.
  • End waiting for material: 96 minutes back, or 4.0 points.

OEE moves by less than that, because it is multiplied by performance and quality. Availability up 4.0 points at 90% performance and 98% quality is about 3.5 points of OEE.

Ending the material wait may need no machine work at all, yet it is worth nearly as much as the failure work. A list that shows only machine faults would have missed it.

A reason is not yet a root cause

The reason is what the operator can see in seconds. The root cause takes longer, and it comes later.

QuestionAnswer
Logged reasonLabeller jam
Why did it jam?Bottles reached the labeller at an angle
Why at an angle?The guide rail was moved at the last changeover
Why was it moved?The changeover standard has no setting for that rail
Example only. The operator logs the first row, and the team finds the rest when the bar reaches the top.

Do not ask operators to name the last row during a stop. Run root cause analysis only for reasons at the top of the chart.

How to get honest reasons from the floor

  • Ask at the moment of the stop. A reason typed at the end of the shift is mostly memory.
  • Name what people can see. "Labeller jam" works, and "mechanical issue" does not.
  • Allow "cause not yet known". Then follow it up the same day, so it does not grow.
  • Never use reasons for blame. If people fear a reason, they pick a safer one.
  • Check against the machine. Compare logged minutes with what a counter or signal shows. See downtime data accuracy.

For the log itself, see operator downtime logging and the free downtime log template.

What to do this week

Log every stop with one reason for five working days. Build two rankings, one by minutes and one by count.

Pick the top reason on each list and name one owner for it. Then look at how your sector tends to lose time, such as in packaging or food and beverage.

See it live in Fabrico

Many reasons are hard to see once the line is running again. Where cameras are fitted, Fabrico can stream video of a stop on request next to its machine data, and the operator tags the cause at the line. See how Fabrico uses cameras for OEE.

What people ask next

What is the most common reason for machine downtime?

We found no neutral source that ranks reasons across plants, and the answer changes by process. Rank your own stops by minutes and by count. The downtime Pareto chart shows how.

What is the difference between a downtime reason and a downtime category?

The reason is what stopped this machine, such as a label web break. The category is the group above it, such as unplanned technical. See machine downtime categories.

How many downtime reasons should operators choose from?

Few enough to pick in seconds, with the list specific to the machine. Merge reasons nobody uses and split one that holds a very large share of minutes. See downtime reason codes.

What should we do when nobody knows why a machine stopped?

Log "cause not yet known" and review it the same day. Do not let it become a permanent "other" bar. The unknown downtime guide covers how to shrink it.

Are short jams a downtime reason?

Yes, if they pass your stop threshold. Below it they show up as speed loss, and the reason gets lost. See micro-stops.

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

← More OEE Lab guides · Build a downtime Pareto chart →

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