This article is for plant managers, production managers, continuous improvement and maintenance leaders who review a downtime Pareto every week and keep finding the same long breakdowns at the top. It explains why short, repeated stops of a few seconds rarely reach the top of a Pareto sorted by minutes, what that does to OEE decisions, and how to read stop data by frequency as well as by duration, so that 30-second stops get the attention their real cost deserves.
- Two machines can lose the same minutes for opposite reasons.
- A Pareto by minutes hides frequent 30-second stops.
- Count, duration and restart losses each tell part of the story.
- Frequent stops need robustness, long ones need faster recovery.
Same downtime, very different machines
Ray Harkins, who teaches through The Manufacturing Academy and is general manager of Lexington Technologies, makes this point in “Downtime Frequency and Duration”, chapter 6 of his book Measuring Manufacturing Effectiveness (2026), also available as an article on Accendo Reliability. Comparing equipment with the same total downtime, one interrupted often and briefly, the other rarely but for long periods, he writes: “Total downtime does not distinguish between these realities. Downtime behavior does.”
His chapter frames the two dimensions in reliability terms: how often interruptions occur relates to MTBF, how long recovery takes relates to MTTR, and availability on its own does not tell you which pattern you are facing. He also observes that frequent short interruptions tend to become normal for operators and, without automated data collection, are recorded inconsistently. And he argues that the two patterns call for different improvement work: reliability and robustness for frequent interruptions, maintainability and recovery capability for long ones. This article takes that framework to the tool most teams use to decide what to fix first: the OEE Pareto.
Why a Pareto sorted by minutes buries 30-second stops
A downtime Pareto is usually built on total minutes lost per reason. That is a sensible default, because minutes convert directly into lost output. But it has three blind spots when stops are short and frequent.
- The logging threshold. If stops are written down by hand, or only counted above a minimum duration, the 30-second stop is never recorded. It does not appear low in the Pareto; it does not appear at all. Its time ends up inside the performance loss, as an unexplained gap between the theoretical and the actual rate.
- Dilution across reasons. Even when short stops are captured, they are often grouped under vague labels (jam, adjustment, sensor) or spread over many reasons. Each bar stays short, while a single breakdown with a clear label and a long duration takes the top spot.
- The cost of a stop is not its duration. The recorded 30 seconds is the time the machine stood still. It does not include what follows: slower cycles after the restart, the first parts to check, the operator pulled away from another task to clear the jam. Harkins describes the same mechanism when he explains how frequent interruptions fragment production and create repeated ramp-up losses.
The result is a Pareto that is accurate about long stops and silent about short ones. Improvement effort then goes to the breakdowns everyone can see, while the line keeps losing capacity in increments too small to notice one at a time.
An illustrative example: one week, three Paretos
Take, for example, a packaging machine over one week of production (the figures below are invented for illustration). The stop data contains four reasons. Sorted three different ways, the same data tells three different stories.
| Stop reason | Number of stops | Total minutes | Average duration |
|---|---|---|---|
| Film jam at sealing | 240 | 120 | 30 s |
| Motor breakdown | 1 | 150 | 150 min |
| Product missing at infeed | 90 | 60 | 40 s |
| Changeover overrun | 3 | 75 | 25 min |
Sorted by total minutes, the motor breakdown comes first (150 minutes) and the film jam second (120 minutes): the natural conclusion is a maintenance problem. Sorted by count, the film jam is first by far with 240 stops, followed by infeed starvation: the conclusion becomes a process problem. Sorted by average duration, the breakdown and the changeover overrun dominate, which points to recovery time. None of these views is wrong. Each one answers a different question.
Now add the restart. If, in this example, each jam also costs around 20 seconds of slower cycles before the machine is back at nominal rate, the 240 jams cost another 80 minutes that show up as performance loss, not as downtime. The film jam then becomes the largest loss on the machine, ahead of the breakdown, although it never reached the top of the duration Pareto. And if the manual log only records stops of five minutes or more, the jam and infeed rows disappear entirely: the Pareto shows a breakdown and a changeover, and the team never sees 330 of the 334 stops of the week.
Frequency and duration: a matrix to read next to the Pareto
A simple way to keep both dimensions in view is to place each stop reason on two axes: number of stops over the period, and average duration per stop. Four zones appear, each calling for a different kind of action. The table follows the distinction Harkins draws between reliability work for frequent interruptions and maintainability work for long ones.
| Zone | Typical pattern | Position in a duration Pareto | Usual lever |
|---|---|---|---|
| Frequent and short | Jams, misfeeds, sensor resets, momentary faults | Low, or absent if not logged | Process robustness: settings, material, guides, sensors, standard work |
| Rare and long | Breakdowns, major failures | At the top | Maintainability: diagnosis, spare parts, skills, access for maintenance |
| Frequent and long | Chronic problem on a critical function | At the top | Treat first, usually on both reliability and recovery |
| Rare and short | Isolated events | Low | Monitor, no dedicated project |
The frequent and short zone is where micro-stops live. Each stop is too short to trigger an intervention on its own, which is exactly why it keeps coming back. A 30-second jam that happens 40 times a shift is not a nuisance: it is a process that has learned to stop.
How to rebuild an OEE Pareto so micro-stops show up
- Capture every stop, with no minimum duration. Define the micro-stop threshold you will use for reporting, but record all stops, so that the threshold becomes a filter and not a hole in the data.
- Keep reasons specific. Replace catch-all labels such as jam or adjustment with where and what: jam at sealing, product missing at infeed. Short stops can only be ranked if they are not diluted.
- Build two Paretos, not one. One by total minutes, one by number of stops, on the same machine and the same period. Read them side by side in the weekly review.
- Look at the restart. Check the cycle rate in the minutes after short stops. If the machine runs slower after each restart, the real cost of a frequent stop is higher than its recorded duration.
- Place each reason on the frequency and duration matrix and choose the lever according to the zone: robustness for frequent stops, recovery for long ones.
- Track the count as a progress indicator. For a micro-stop project, the number of stops per hour or per shift is often clearer than minutes, because it moves as soon as a cause is removed.
Measuring the stops nobody writes down
All of this rests on one condition: having every stop, short ones included, with its start time and duration. A manual log cannot provide that. Operators do not interrupt their work to note a 30-second jam, and nobody should ask them to. TEEPTRAK measures machine stops, micro-stops and cycle rate automatically, with a module wired in the electrical cabinet and a sensor at the edge of the line, without connecting to the PLC. Installation takes about an hour per machine and OEE is available within 48 hours. Operators only qualify the reason for a stop, so the Pareto by count and the Pareto by minutes come from the same data. TEEPTRAK is used in more than 700 plants in over 30 countries.
Further reading
- Harkins, R. (2026). Downtime Frequency and Duration. In Measuring Manufacturing Effectiveness, ch. 6, pp. 29 to 34. Amazon KDP. Also available as an article on Accendo Reliability.
- Eliminating Micro-Stops: Pareto, Top 3 Root Causes, and Targeted SMED Project
- MTBF MTTR unplanned downtime: measuring equipment reliability
FAQ
What duration counts as a micro-stop? There is no universal threshold. Many plants set a limit of a few minutes or less. What matters is to define one threshold per machine family, apply it the same way on every shift, and still record all stops, so the threshold can be adjusted without losing history.
Do micro-stops belong to availability or to performance in OEE? It depends on the rule your plant sets. Some sites count every stop as availability loss; others class stops below the threshold as minor stops within performance loss. Both are defensible if the rule is written down and applied consistently. A Pareto by count works in either case.
Should we stop using the Pareto by minutes? No. Minutes remain the right measure of lost capacity, and long stops deserve their place at the top. The Pareto by count complements it by showing the frequent stops that minutes hide, and the restart losses that follow them.
0 Comments