OEE is the result of: Availability % x Performance % x Quality %
Availability is the percentage of time that a machine is available to produce during the measured period.
For example, if a shift is 7 AM to 3 PM with a half hour lunch and two fifteen minute breaks, then there are 7 hours, or 420 minutes, available during that shift. If the machine was actually running 315 minutes during this shift (downtime of 105 minutes) then availability is 75%. The machine was available to produce 75% of the time.
Next is performance. First a target time needs to be decided on. This should be how fast a single part can be produced on the machine based on the largest limiting factor of the process.
For example, if a plastic part is being extruded, the part is 3 pounds per piece and the extruder can only pass 3 pounds per minute at maximum speed, then the target would be 1 minute per piece. So, taking our first example of availability where we ran for 315 minutes during the shift, for a rating of 100% performance we would need to produce 315 parts at 1 minute a piece during this time period. Lets say we produced 300 parts. Then 300 divided by 315 gives us a 95% performance rate.
Last, there is quality. The quality percentage is how many acceptable parts were produced out of the total produced on the machine. Again, continuing with the previous examples, if every one of the 300 parts we produced was acceptable then we would get a 100% quality rate.
For this example, let’s say that only 270 parts were acceptable (30 parts rejected). If we divide 270 acceptable parts by the 300 total parts we get a 90% quality rate.
So now it's time to calculate the overall effectiveness of our machine. We had 75% availability, 95% performance rate, and 90% quality rate. Therefore, this machine has an OEE percentage of 64% (75% x 95% x 90%.)
Sounds low, doesn't it? The machine performed at a decent rate, quality was 9 out of 10, and we were available most of the day but together it tells a different story. We only produced two-thirds of the parts on this machine as we expected to and we also had some waste due to quality problems. Now we have something to work on and improve.
Keep track of all data for the long term. Compare month to month, shift to shift and so on. Does one shift perform consistently better than another? Maybe they do something that the other shifts don't do, such as a different tooling change or better trained operators. Does the machine perform notably better in the summer than the winter? We'll maybe it would then be better to take the two week maintenance outage during the winter months to make yearly repairs on the machine instead of in the summer. In the end, tracking OEE on a machine is not a tool that is going to make a machine produce more or run better, but is instead a way to determine which of your tools are working and which ones are not.