Ideal run rate = 60 ÷ tool cycle time in seconds x units per cycle, which gives units per minute at the highest speed the machine is accredited to run for that product. On a line, use the rate of the bottleneck. If you have no accredited figure, start from your best 4 hours in a row out of the last 400 running hours.
Which route should you take?
You rarely have to invent this number, because it can be worked out from what you already have. Pick the first route below that you can fill in.
One machine with a set tool cycle. Calculate units per minute from it.
Find the rated speed of each machine, in the same unit. The lowest one sets the rate.
Take the supplier figure for that format, and prove it on the floor before you use it.
Use your best 4 hours in a row from the last 400 running hours as a starting rate.
The rules below follow Overall Equipment Effectiveness (R. C. Hansen, 2001). The rate is a per-unit number, so count the same unit you count in OEE.
Route 1: from cycle time and units per cycle
Example: a blister press closes a tool every 4 seconds, and each stroke makes 12 blister cards. It is accredited at that cycle.
If one cavity is blocked for a run, keep the rate at 12 cards. The empty cavity is a real loss, and lowering the rate would hide it.
Route 2: from the bottleneck of a line
On a line, the ideal run rate is the rate of the machine that limits the line, not the fastest one. Example: a tray line with four machines and a rate per machine in packs per minute.
Watch the units. The case packer here is rated at 5 cases per minute with 10 packs in a case, and 5 x 10 = 50 packs per minute.
For one machine that is not the bottleneck, the ideal rate is the rate needed to feed the bottleneck. A tray former measured on its own at 70 still only needs 45 here, so it simply caps at 100%, and line OEE vs machine OEE covers the difference.
Not sure which machine limits your line? The bottleneck analysis guide shows how to find it, and the V-graph shows how the others should be set around it.
Route 3: from a rated speed
A brochure speed is often for one format under one set of conditions. Ask for the rate at your size and material, then test it.
Run the machine at that rate for a few hours with normal material and a normal crew. If it holds and the parts are good, write it down as the accredited rate.
If it cannot hold, do not simply take the speed you got instead. Note what stopped it, because that is what the rate gap is measuring.
Route 4: from your best 4 hours of 400
Hansen suggests this as a starting point when no figure exists. Take the last 400 running hours and find the best 4 hours in a row.
Add those four hours and divide by the minutes. Here, 5,880 + 6,060 + 6,000 + 6,060 = 24,000 units, and 24,000 ÷ 240 minutes = 100 units per minute.
In Excel, put hourly counts in column B from row 2. Paste =SUM(B2:B5)/240 into C5 and drag it down to row 401, which fits 400 hours in rows 2 to 401.
Then put =MAX(C5:C401) in a cell outside column C, such as E2. That is the rate in units per minute.
This gives a demonstrated rate, so treat it as a floor for the true ideal. Using a rolling window also filters out a single lucky hour.
One rate per family, kept in a table
Do not average rates across products. Group products of the same size and format, and give each group a rate.
| Family | Tool cycle | Units per cycle | Ideal rate | Per hour | Ideal cycle |
|---|---|---|---|---|---|
| 250 ml | 4 s | 12 | 180 / min | 10,800 | 0.333 s |
| 500 ml | 4 s | 8 | 120 / min | 7,200 | 0.500 s |
| 1 litre | 4 s | 6 | 90 / min | 5,400 | 0.667 s |
Process lines work the same way in other units. An extruder accredited at 540 kg per hour runs at 9 kg per minute, and that is the rate if you count output in kg.
If one product in a family runs hard, keep the family rate and note why. That keeps the difficult product visible, and the guide to OEE by product shows how to read it.
Put it to work in the OEE sum
Example: the 100-per-minute line from route 4 has 480 scheduled minutes and 80 minutes of stops, so it runs for 400 minutes. It makes 35,000 units, and 34,300 are good.
| Step | Calculation | Result |
|---|---|---|
| Actual run rate | 35,000 / 400 | 87.5 / min |
| Performance | 87.5 / 100 | 87.5% |
| Availability | 400 / 480 | 83.3% |
| Quality | 34,300 / 35,000 | 98.0% |
| OEE | (34,300 / 100) / 480 | 71.5% |
The full method for that factor is in how to calculate OEE performance. Once the rate is fixed, the speed loss calculation turns the gap into minutes.
Common mistakes
- Using last month's average. The line then scores near 100% on performance and its speed loss disappears from view.
- Mixing strokes and units. Check that the count and the rate use the same unit.
- Using the fastest machine's rate for the line. A rate no machine chain can hold makes every shift look poor.
- Changing the rate after a bad week. OEE trends then stop meaning anything. Change it only after a real change on the machine, and record the date.
- Getting over 100%. That means the rate is too low or the counter is wrong, as performance over 100% explains.
The best 4 hours are easy to find once counts are logged by the hour. Fabrico takes speed and counts from PLCs or IoT sensors, including external sensors on older machines. See how Fabrico reads machine counts.
What people ask next
What is the difference between ideal run rate and ideal cycle time?
They are the same number turned upside down. Ideal rate = 60 ÷ ideal cycle time in seconds, so 1.5 s per unit is 40 units per minute. The cycle time calculator gives your real time per unit, to compare with it.
Is ideal run rate the same as nameplate speed?
Not always. Nameplate is the supplier's figure, often for one format, while the ideal rate is the highest speed accredited for your product. Test a nameplate figure before you use it, and use the bottleneck on a line.
Should ideal run rate include breaks or small stops?
No. It is the speed while running, with no allowance for stops, breaks or small jams. Those losses are measured by availability and performance, so building them into the rate would hide them.
What if a machine beats its ideal run rate?
Performance goes above 100%, which means the rate is too low or the counter is wrong. Check the counter first, then re-accredit the rate if the speed is real and the parts are good. See performance over 100%.
How often should I change the ideal run rate?
Only when the machine, tool or process really changes, and after the new speed has been proven. Record the old rate, the new rate and the date, so a step in the OEE trend has an explanation.
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.