OEE calculation: the formula, three worked examples and the mistakes to avoid

The OEE formula in one line

Overall equipment effectiveness (OEE) is the share of planned production time that was truly productive: the machine running, running at its ideal speed and making good parts. It is the product of three factors.

The shortcut works because the three factors cancel: run time and total count drop out, leaving the time it should have taken to make the good parts divided by the time you planned to run. Use the factors when you want to know where the loss is, and the shortcut as a check on your arithmetic. The OEE calculator does both from the same six numbers.

Availability, performance and quality, one at a time

Availability asks how much of the planned time the machine actually ran. Breakdowns, changeovers and waiting for material all take it down. Performance asks how fast it ran while it was running, so slow cycles and the many short stops that never reach a stop log both land here. Quality asks how many of the parts made were good the first time.

That split is why OEE is useful beyond the single number. The six big losses sort the same way: equipment failures and setup and adjustment are availability losses, small stops and reduced speed are performance losses, and start-up rejects and production rejects are quality losses. A line at 70% OEE with availability as the weak factor needs a different team from one with performance as the weak factor. The six big losses worksheet puts each loss under its factor.

Example 2: a fast line measured in units per minute

Fast lines are usually rated in units per minute, not seconds per unit, so convert first. This example is illustrative. A filling line has an ideal rate of 120 units a minute, which is an ideal cycle time of 0.5 seconds. The shift is 480 minutes with 45 minutes of planned stops. Unplanned downtime was 52 minutes: a 17-minute breakdown, a 25-minute changeover and 10 minutes waiting for empty containers. The line counted 39,100 units and 780 failed the quality check.

The weak factor is performance, not availability. In a real plant the first question would be what slowed the line down while it ran: short stops that never made the stop log are the usual answer, which is what the micro-stops page is about.

OEE for a line or a plant

To get OEE for several machines, a line or a plant, do not average their percentages. Weight each machine's OEE by its planned production time, or add up the ideal time of all the good output and divide by all the planned time. The two methods give the same answer.

An illustrative contrast: Machine 1 was planned for 8 hours and reached 80% OEE. Machine 2 was planned for only 2 hours that day and reached 50%. The simple average of the two percentages is 65%. Weighted by planned time it is (8 × 0.80 + 2 × 0.50) ÷ 10 = 74%. The machine that ran most of the day should count for most of the result, and the average gives the short run the same say as the long one.

The same trap catches dashboards. A plant figure that is typed in separately, or averaged from line figures, will drift from one built from its inputs. Roll up from the underlying times and counts and keep the definitions the same on every line; the KPI dashboards page makes the same point about ratio KPIs such as OEE, which should roll up from their inputs and never as an average of percentages.

Eight calculation mistakes

Most of these make OEE look better than the floor feels, which is a useful test: if the number is high and the team says the machine is always stopping, check the stop log and the ideal cycle time first. For short stops, the micro-stops page explains why they are usually the missing loss.

From number to action

The OEE figure tells you how big the gap is. The three factors tell you which kind of loss to look for, and the stop log tells you which cause to start with.

• Availability is the lowest factor: sort the stop log by minutes, by number of stops and by cost, and look at changeovers separately from breakdowns. Reducing machine downtime works a month of stops through that method.
• Performance is the lowest factor: start with short stops and slow cycles. The micro-stops page describes logging each one in one tap and adding them up by cause with a cost.
• Quality is the lowest factor: look at what is being rejected and where it is found. The loss and cost management page shows how LeanSuite puts a cost on each loss so the biggest is fixed first.
• A fix that needs an owner and a date belongs in an issue list, not in the OEE report; daily issue management is where LeanSuite keeps those.

A plain statement of scope: LeanSuite does not read OEE from machines. The figure comes from counts and times that people record, and the KPI dashboards track OEE next to downtime by reason and output against plan. If you want OEE captured automatically from machine signals, that is machine-monitoring software, which LeanSuite is not; the compare page is a starting point for tools that work that way.

Definitions and sources

• OEE as availability × performance × quality, and the shortcut of good count × ideal cycle time ÷ planned production time: as on the OEE calculator page. Planned production time is the shift minus planned stops; run time is planned production time minus unplanned downtime (OEE.com (Vorne), "Calculating OEE").
• Six big losses: from Seiichi Nakajima's TPM work (Introduction to TPM, Productivity Press, 1988), read through OEE.com (Vorne), "Six Big Losses". Equipment failures and setup and adjustment are availability losses; small stops and reduced speed are performance losses; start-up rejects and production rejects are quality losses.
• The 85% world-class OEE (90% × 95% × 99%) is attributed to Nakajima, Introduction to TPM, and is quoted on OEE.com (Vorne), "World-Class OEE".
• TEEP is OEE × utilization, with utilization planned production time ÷ calendar time, as on the TEEP calculator page. OLE has no single standard, as noted on the OLE calculator page.
• All numbers in the three worked examples and the roll-up are made up for illustration. They are not from a real plant.