TL;DR Overall Equipment Effectiveness (OEE) is a useful metric to understand the relative health of a process, but more granularity is needed for feedback on where improvements are possible. We recommend tracking performance and quality by product, and explicitly defining what constitutes planned versus unplanned downtime.
How is the line running?
Are we winning the day?
Am I doing better, worse, or staying the same?
OEE (overall equipment effectiveness) is one way to answer those questions.
Like any metric, OEE is only as useful as its inputs and in turn, can be misleading. Used properly, it can both provide a pulse of the process in question and with its intermediate calculations, insight into where it can be improved.
There are three main components to OEE:
Availability is defined as the total running time divided by the total planned running time. In more common terms, how much is the line running versus how much it was planned to run.
This takes into account and excludes planned downtime or idle time, but the definition of planned/unplanned is up to the user. For a useful measurement, we recommend automating the recording of downtime to improve the accuracy of this measurement.
On a performance basis, this is asking the question
how close to ideal speed is my process running?
There is an expected speed that the line would run based on the current product and operating conditions. If an operator slows the line to fill a hopper or adjust a machine, it decreases performance. If the line is run over the ideal speed, it increases performance.
The most important piece of a performance calculation is an appropriately defined ideal speed. This could be the nameplate speed of the slowest piece of machinery in the process, but it’s recommended to use historical performance data to arrive at ideal speeds by product/conditions.
Performance is calculated as
(volume produced/time running)/ideal line speed
Performance is easier to automatically calculate than Availability, but as noted above, ideal speed should be well considered.
The final component of OEE is quality, defined as how many units are produced at the end of the process and able to be sold versus how many started.
It is recommended to measure quantity a few steps through the process in order to provide granularity on the source of the losses in quality.
Multiply availability, performance, and quality together to generate the line’s OEE.
Trended over time, OEE is an effective way to assess the health of a process or line, but the individual constituents must also be trended to understand why OEE increased or decreased. This supporting information gives the opportunity to ask
What caused the decrease in availability, performance, or quality?
Can we do something differently to improve what decreased?
Because isn’t that the point of measuring effectiveness in the first place?