Speed loss, minor stops, ideal cycle time and micro-stop observation — explained for food and FMCG manufacturing with real worked examples.
Performance is the second of the three OEE components. It measures how fast your equipment is running compared to its maximum possible speed during the time it is actually running. If Availability tells you how much of your scheduled time the line was running, Performance tells you how well it was running during that time.
A Performance of 100% would mean the line produced exactly as many units as it theoretically could have at full speed for every minute it was running. In practice, no line achieves 100% Performance — there are always minor stops, small speed reductions, and brief interruptions that shave output below the theoretical maximum.
Performance is caused by two specific loss types from the Six Big Losses framework: reduced speed losses (the line running slower than ideal without stopping) and minor stops (brief, frequent interruptions that are not recorded as downtime). Both reduce actual output below theoretical maximum.
The 14,680 packs difference between theoretical maximum (31,200) and actual output (26,520) represents the combined Performance loss — some from running slower than 80 packs/minute and some from minor stops not captured in the downtime log. In practice, breaking this down requires a micro-stop observation study.
Ideal cycle time is the theoretical minimum time required to produce one unit under perfect conditions at maximum rated speed. It is the denominator that defines what "100% Performance" looks like for your specific equipment.
Getting ideal cycle time right is critical. If you set it too slow (i.e. use actual average speed rather than the machine's rated maximum), your Performance figure will appear artificially high — you will think you are performing well when you are simply measuring against a lower bar.
The simplest and most defensible source. Every piece of equipment has a rated speed. Use this as the ideal unless there is a documented engineering reason not to. If the machine was commissioned at a lower speed, question whether that lower speed was ever challenged.
For older equipment without clear nameplate data, observe the line running under the best conditions you have achieved in the last 6 months. The fastest sustained speed observed over a 10-minute period — without quality compromise — becomes your ideal cycle time baseline.
If your line averages 72 packs/minute and you set ideal cycle time at 72 packs/minute, your Performance will always be close to 100% — not because you are performing well, but because you have set the target to match your average output. This destroys the value of OEE entirely.
Once set, ideal cycle time should not change unless the equipment changes. Adjusting it downward when Performance looks poor is the manufacturing equivalent of moving the goalposts. It undermines trust in the data and makes improvement invisible.
Performance is damaged by exactly two loss types. Understanding the difference is important because they have different causes and require different solutions.
The line is running continuously but slower than its ideal speed. Output is reduced, but there are no stops to log. This is a sustained, continuous reduction in throughput.
Brief, unplanned stoppages — typically under 5 minutes — cleared by the operator without maintenance involvement. They are frequent and intermittent rather than continuous, and usually go entirely unrecorded.
Worn chains, drive belts, bearings, and cam followers all degrade speed capability over time. If a line that once ran at 80 packs/minute now runs reliably at 72, the delta is likely mechanical degradation. A condition audit will identify the components responsible.
Running slower to stay within fill weight tolerance, sealing temperature windows, or label placement spec is a real and valid reason to reduce speed. But it should be documented as an engineering constraint — not an informal operator decision. And the root quality issue should be solved, not managed by running slower forever.
Many lines were commissioned at a safe speed and never challenged. The commissioning engineer set 70 packs/minute because it was stable on day one. Five years later the line could reliably run at 78 packs/minute but nobody has tested it because the old number is in the system. A structured speed trial changes this.
Speed loss is visible when you compare actual output to theoretical maximum output for a period when the line was running continuously (no stops). If the line ran for 60 minutes without a recorded stop and produced 4,200 units against an ideal of 4,800, the speed loss is 600 units — 12.5% — even though nothing was logged.
If you don't know what the line can do at its best, you can't measure how far short of it you are. Run speed trials for your top 5 SKUs and document the results.
Walk the line with a maintenance engineer and inspect the components most likely to degrade speed: drive chains, belts, bearings, cams, and infeed/outfeed conveyors. Replace worn components and rerun the speed trial.
If the line is running slower because of a quality concern, fix the quality root cause rather than accepting a permanent speed reduction. Running at 72 instead of 80 to avoid fill weight drift is a symptom of a filling system problem, not a speed setting problem.
Once validated ideal speeds are confirmed, update the system settings. Operators should not be able to informally reduce speed without a documented reason. If speed changes are needed, they should be recorded and reviewed at the daily performance meeting.
Minor stops are the single most underestimated source of OEE loss in food and FMCG manufacturing. They are too short to log as downtime, too frequent to feel significant in the moment, and entirely invisible in most reporting systems. Yet their cumulative impact can be larger than all recorded downtime combined.
A micro-stop observation study is the only reliable way to quantify and categorise minor stops. It requires a person — an engineer, CI coordinator, or trained operator — to stand at the line for a full shift and log every stop manually, regardless of duration.
Sites that estimate 10–15 minor stops per shift typically find 40–70 when they actually count them. The human brain filters out brief interruptions that feel routine.
A Pareto of stop locations and failure modes almost always shows a small number of recurring issues responsible for the vast majority of lost time. Fix those first.
In food lines, the highest concentration of minor stops is typically at product transfer points — infeed conveyors, turn tables, pack accumulators, and case packer infeeds — where product spacing and alignment are most sensitive.
The study often reveals stops of 8–15 minutes that operators cleared without logging. These are not minor stops — they are unreported downtime events that should be in the Availability data.
One shift is not enough. Run the study on different operators, different shifts, and ideally on different days of the week. Patterns only become clear with sufficient data.
Rank by total time lost, not by count. A stop that happens 50 times but lasts 15 seconds each is less important than one that happens 10 times but takes 3 minutes to clear.
Procedures do not fix minor stops. Guides telling operators to clear jams more carefully do not reduce the jam rate. The physical cause — a worn guide rail, a misaligned sensor, a product flow issue — needs an engineering fix.
After the fix, rerun the observation study on the same shift and compare the before/after Pareto. If the fix worked, that stop category disappears from the top 3. Move to the next cause.
Once the major physical causes are fixed, install a simple tally system for operators to log stops over 2 minutes. This bridges the gap between micro-stop observation (resource-intensive) and the daily downtime report (only captures longer stops).
Performance benchmarks vary significantly by product type, packaging format, and line complexity. These ranges reflect typical observed Performance in UK and European food and FMCG manufacturing.
| Sector | Typical Performance range | Dominant loss type | Key improvement lever |
|---|---|---|---|
| Ready meals / chilled food | 75–85% | Minor stops at sealing and portioning | Sealing head and film tension engineering fixes |
| Bakery (bread, morning goods) | 82–90% | Speed loss on depositing and moulding | Speed trial + dough consistency standardisation |
| Beverage (carbonated / still) | 85–92% | Minor stops on filling heads and capping | Filling head maintenance schedule + jetting frequency |
| Fresh produce (salad, veg) | 70–82% | Speed loss on weighers and minor stops on baggers | Weigher calibration + bagger film guide alignment |
| Baby food (pouches / jars) | 78–87% | Minor stops at filling and capping; speed loss during sterilisation | Filler nozzle maintenance + cap torque engineering review |
| Confectionery / snacks | 80–90% | Speed loss on enrobing and minor stops on wrapping | Enrober temperature management + wrapper timing adjustment |
| General FMCG packaging | 77–88% | Minor stops at case packing and palletising | Case packer infeed alignment + robot gripper maintenance |
| Performance score | Classification | What it typically means |
|---|---|---|
| 95%+ | World-class | Minor stops nearly eliminated; running at or close to ideal speed |
| 88–94% | Good | Occasional minor stops; speed close to ideal but some loss evident |
| 75–87% | Typical food/FMCG | Regular minor stops; speed loss present but partially managed |
| 65–74% | Below average | Frequent minor stops or significant speed loss — investigation required |
| Below 65% | Poor | Major Performance issue — micro-stop study and speed audit needed urgently |
Enter your run time, ideal speed and actual output to get your Performance score instantly.