Everyone in the plant can recite the OEE formula from memory: availability times performance times quality. Ask two people to calculate the same shift, and the numbers almost never match.
OEE, overall equipment effectiveness, measures what fraction of your planned production time turns into good parts running at the ideal rate. You calculate it by multiplying three fractions, availability, performance and quality, and the result runs from zero to one hundred percent.
The formula is not where things break. What breaks it is three words every plant defines differently: what counts as planned time, what counts as downtime, and what counts as the ideal rate. If those three definitions are not written down somewhere, two people calculate the same shift and walk away with two different, equally defensible, equally useless numbers.
The three fractions, one at a time
Availability is run time over planned time. You subtract downtime, planned and unplanned, from planned time, and what is left is run time.
Performance is parts produced over the parts you theoretically could have made. Take the run time and divide it by the ideal cycle time, and that tells you how many parts you should have gotten running flat out. What you actually made, divided by that number, is your performance.
Quality is good parts over parts produced. Of everything that came off the line, how much passed without rework or rejection.
The word that causes the most argument is ideal. The nameplate speed on the equipment and the best shift anybody remembers both make poor stand ins for it. Ideal cycle time is the demonstrated, sustainable rate you can back up with a record, not with memory. Use nameplate speed and your performance looks low on a healthy line. Use the best shift ever recorded and your performance never reaches one hundred.
One shift, with numbers
Take an eight hour shift as an example, 480 minutes. Thirty minutes come out for lunch, so planned production time is 450 minutes. During the shift there were 45 minutes of unplanned downtime, an equipment failure and a material changeover. Run time is 450 minus 45, 405 minutes.
Availability = 405 over 450 = 90 percent.
The ideal cycle time on this line is one minute per part. At that rate, 405 minutes of run time could have produced 405 parts. The line actually made 360. Performance = 360 over 405 = 88.9 percent.
Of those 360 parts, 18 came back as rejects. Good parts: 342. Quality = 342 over 360 = 95 percent.
OEE = 90% × 88.9% × 95% = 76%.
The check I run before I report the number
There is a short way to confirm you did not lose the thread somewhere along the way. OEE is also equal to good parts multiplied by ideal cycle time, divided by planned time. In this example, 342 good parts times one minute of ideal cycle time, divided by 450 planned minutes, comes out to 76 percent. If that number does not match the one you got multiplying the three fractions, the mistake is in one of the three, not in the multiplication. I run this check every time before an OEE goes into a report.
Where the points actually go, and what to do about each one
We have covered how to calculate it. Now comes the part that actually matters, which is where the points are leaking, because each fraction gets attacked differently.
- Low availability: start with the line between planned and unplanned downtime. That is where production and maintenance argue the most about whose number is right.
- Low performance: it is rarely one big failure. It is usually micro stops under a minute that nobody logs, plus reduced speed that becomes habit.
- Low quality: watch for rework that gets counted as a good part on the production count but does not pass the customer’s standard. That inflates the number without the problem actually being solved.
Why this number reaches past maintenance
OEE almost always gets used to justify capital spending on equipment, and that is exactly where the number matters more than it looks like it does. If your availability looks bad because downtime gets misclassified, you are asking for budget to fix the wrong problem. Before proposing new equipment, it is worth confirming which fraction is actually losing you the points.
Start here
Take yesterday’s shift. Pull the three fractions apart, with their definitions written down, before you multiply them. Run the check, good parts times ideal cycle time, divided by planned time. If the two numbers do not match, you already know where to look. If you want to work through it together on real data from your line and decide what to attack first, that is part of what I work on in problem solving and continuous improvement.
What is the formula for OEE?
OEE = availability × performance × quality. Availability is run time over planned time, performance is parts produced over the theoretical maximum at the ideal rate, and quality is good parts over parts produced.
What counts as a good OEE?
There is no universal cutoff that holds up across every process and every piece of equipment. What does hold up is comparing the same shift against itself over time, using the same definitions of planned time, downtime and ideal rate.
Why does my OEE not match someone else’s calculation of the same shift?
Almost always because planned time, downtime or ideal rate are not defined the same way for both of you. Write the three definitions down once, and the number stops moving between reports.