This article is for plant managers, production managers, industrial engineers and continuous improvement leads who calculate OEE and have to set, or defend, the reference speed of each machine. It deals with a question that is often settled once when tracking starts and then forgotten: which ideal cycle time to use between the manufacturer’s rated speed, the best observed cycle and the standard rate from routings, and how to avoid a poor choice distorting the performance rate for years.
- The performance rate depends entirely on the chosen reference speed.
- Manufacturer, best sustained cycle or routing standard: three figures, three OEE values.
- A reference that is too low hides slowdowns and micro stops.
- One reference per product, revised on facts and logged.
Why the reference speed decides the performance rate
The performance rate compares what the machine produced during its run time with what it should have produced at the reference speed. In other words: ideal cycle time multiplied by the number of parts, divided by run time. Availability and quality are measured on facts (the machine runs or not, the part is good or not). Performance depends entirely on a figure someone chose. Changing the reference changes OEE without anything changing on the shop floor.
That is why two plants in the same group can show very different OEE on identical machines: they do not start from the same reference. Before comparing, you need to know where the figure comes from.
Three possible sources, three different effects
| Source of the speed | Advantage | Risk | Effect on performance |
|---|---|---|---|
| Manufacturer’s rated speed (nameplate, manual) | External, stable, hard to dispute | Often obtained in ideal conditions, sometimes unreachable with your products | Permanently low performance, speed losses overstated |
| Best cycle observed and sustained over time | Reflects what the machine can do with your products | Needs measured data, not a one off record | Realistic performance, usable gaps |
| Standard rate from routings or costing | Already used by planning and cost accounting | Often includes margins and historic slowdowns | Flattering performance, sometimes above 100% |
None of the three is wrong in itself: each serves a purpose. The standard rate is for planning and costing, the manufacturer’s speed is a technical ceiling. For OEE, whose job is to reveal losses, the most useful reference is usually the best cycle actually sustained on your products, with the manufacturer’s speed as an upper control limit.
What happens when the reference speed is too low
A reference that is too low, meaning an ideal cycle time that is too long, is the most common trap because it suits everyone. Its effects:
- Slowdowns disappear. A machine running at reduced speed since a setting change two years ago looks like it is running at its normal pace.
- Micro stops are absorbed. The margin between the reference and actual speed offsets part of the small stops, which no longer show up in performance.
- Performance exceeds 100%. A sure sign that the reference is wrong, and one that undermines the indicator in the eyes of the teams.
- Gains become invisible. A project that speeds the machine up does not move OEE, since the margin already absorbed the gap.
- Capacity is underestimated. An investment or an extra shift gets considered while the machine still has headroom.
For example, take a packaging line (fictitious figures, for illustration only). The manual states 120 units per minute, the best pace sustained for an hour on the main product is 100, and the routing uses 80. While running, the line averages 76. Depending on the reference, the performance rate is 63%, 76% or 95%. Same line, same day, three readings. At 95%, nobody looks for a speed loss. At 76%, there is a gap of 24 units per minute to explain.
One reference per product, not one per machine
A machine that runs several products does not have one speed, it has several. A larger format, a harder material or a tighter tolerance requires a different cycle. Using a single speed per machine has two effects: performance drops when the mix shifts toward slow products and rises when it shifts back to fast ones, without the shop floor changing anything. The right level is the product reference, or at least the product family sharing the same cycle. Each work order then carries its own speed, and performance is calculated part by part.
When and how to revise the reference speed
- Revise after a real change: machine modification, new tooling, new product, new material. This is the most legitimate trigger.
- Revise upward when the data shows it: if the sustained pace regularly beats the reference, the reference is too low.
- Do not revise downward to improve the figure: a bad week is not a reason. Lowering a reference must be justified by a documented technical or quality constraint.
- Date and log every version: who changed what, when and why. Without history, you cannot compare OEE before and after.
- Give the reference an owner: usually industrial engineering, validated by production. It is not for the operator or finance alone to set.
An annual review of references, machine by machine, is enough in most plants, provided triggers are handled as they occur.
TEEPTRAK automatically measures actual speeds, stops and micro stops on every machine, with a module wired into the electrical cabinet and a sensor at the line, with no connection to the PLC. Installation takes about one hour per machine and OEE is available within 48 hours, which quickly shows the pace actually sustained per product to set or check the reference. The solution runs in more than 700 plants in more than 30 countries.
Further reading
- Calculating OEE: Key Elements to Consider and Exceptions
- Injection molding presses: tracking downtime and cycle time drift in real time
- OEE Data Reliability: Common Measurement Errors and Solutions
FAQ
Should the manufacturer’s rated speed be the OEE reference? It is mainly a technical ceiling. As a calculation reference it is often unreachable with your products and gives permanently low performance. The best pace actually sustained on your products is usually more useful.
Is performance above 100% possible? Not if the reference is right. A value above 100% almost always means the reference speed is too low or parts are miscounted.
How often should reference speeds be reviewed? At every real change of machine, tooling, product or material, and during an annual review. Each revision should be dated and justified.
0 Comments