OEE FAQ

Do you define everything in units of parts or units of time?

The answer depends on what's most intuitive for you and your operators. Down Time Loss is usually described in units of hours or minutes. Speed Loss is most often defined in terms such as pieces per hour (rate) or seconds per piece (cycle time). Quality Loss is typically looked at as a ratio of rejects to total production. Other quantity related measurements like pounds, kilos or meters may also be used. All of these eventually are converted (normalized) to a percentage that shows actual vs. potential.

Is OEE data subject to misinterpretation?

Without a doubt! The two shift example in the Calculating OEE section is a great example. OEE tells you nothing about how much your resources actually cost in dollars, what the easiest improvement actually is, or how much it will cost you to make that improvement. What you should look for in OEE is losses and bottlenecks that can be eliminated for a cost/benefit that makes sense. OEE data is only meaningful in the context of your situation and your efforts to improve it.

How do you determine the Ideal Cycle Time or Ideal Run Rate of a machine or process?

The simple answer is the Nameplate Capacity - the stated capacity given by the manufacturer or engineer. But like the speedometer in your car, this number may only be an approximation and may vary considerably with machine age and other circumstances. Ideally you should experiment, take measurements and come up with your own number.

How do you define whether a machine is "producing"?

If the machine is producing anything at all, the production clock is running. Even if a machine is not producing, if production is planned, the production clock is running.

How do you minimize number crunching, spreadsheets and reports and make OEE truly simple?

Data isn't very useful if you don't see it until tomorrow, or at next week's production meeting. The printout in the manager's office isn't nearly as helpful as having real-time data, with visual cues, already digested into a form that is useful to you and your operators. Automate the OEE data collection and calculation process, and spend your time dealing with the issues that impact productivity here and now.

What period of time do you use for an OEE calculation?

You can choose any period you want. Most commonly it's based on the working cycle of the factory, i.e. one shift (8 hours or 480 minutes). Other time frames include one day, one job, or even continuous monitoring.

Is it possible to have processes that exceed 100%?

No. If you obtain readings greater than 100%, then at some point of the process you are not defining something correctly. It means that you were underestimating the capacity of a process when you input your parameters. This is most often an issue with Performance, when the Ideal Cycle Time or Ideal Run Rate have not been correctly characterized.

If you have a series of machines or stages in a cell with different capacities, which stage do you base your Ideal Cycle Time (or Ideal Run Rate) on?

Typically, when a cell is “balanced”, meaning that the Nameplate Capacity of all contributing machines, stations, or stages is designed to produce at reasonably close to the same rate, the fastest stage (not the slowest), should be your target. This forces you to recognize all other bottlenecks and strive to improve them.

In cases where a cell is not balanced, some analysis should be performed. Probably the best way to begin is to look at the cell in aggregate to determine if any of the associated stages have a Nameplate Capacity that far exceeds the others. These stages are typically known as “accumulator/buffers”, or may be the result of a piece of equipment not originally designed for use in the particular cell but later added, and should be disregarded when calculating the Ideal Cycle Time for the cell.

How do I calculate OEE for my entire plant?

This is a very interesting question that touches on a tricky and somewhat controversial topic (as discussed below). But first, here are two reasonable options for calculating Plant OEE:

• Calculate Plant OEE Using a Straight Average
• Calculate Plant OEE Using a Weighted Average

Calculate Plant OEE Using a Straight Average – The simplest method of calculating Plant OEE is to average the OEE scores of all production assets (i.e. add together the OEE scores of all production assets and then divide that result by the number of production assets).

In a plant with three production assets the calculation would be:

(OEE₁ + OEE₂ + OEE₃) / 3

Calculate Plant OEE Using a Weighted Average – An improved (albeit more complex) method of calculating Plant OEE is to calculate a weighted average of the OEE scores of all production assets. A weighted average “weights” the OEE score of each production asset to take into account its relative importance. The weight can be any factor that assigns relative importance; however, it is recommended to assign weights based on the value added by the production asset. Doing so, helps keep the focus on improvements that add the most value – to the bottom line.

In a plant with three production assets, the calculation would be:

((OEE₁ x Weight₁) + (OEE₂ x Weight₂) + (OEE₃ x Weight₃)) / (Weight₁ + Weight₂ + Weight₃)

Calculating Plant OEE can be very useful for monitoring trends (such as whether a given plant is improving OEE over time) or as a rough gauge of where a given plant lies in the OEE benchmarking spectrum. However, one should be very cautious about using OEE to compare different plants, or even to compare different production assets. Unless production assets are running identical products on identical equipment under identical conditions, comparing their OEE metrics is somewhat akin to comparing apples to oranges.

Do OEE benchmarks, i.e. Theoretical Capacity, ever change?

Yes, but only if the process is improved in some tangible and measurable way.

Can OEE become a political football?

Yes, to the extent that it is misinterpreted and to the extent that it is used for political purposes rather than to genuinely work towards sustainable improved productivity. Of course that's true of all numbers and measurements. The important thing is to keep the core objective in mind - to produce more quality product with less waste and fewer headaches. Therefore you should be very clear about exactly what you want and make sure you are measuring it. Then you can create incentives for operators to deliver that result.

If there are several interpretations for the cause of a problem, where do you start?

Always look for the simplest and most direct explanation (the root cause). Obviously there are exceptions, but here's what's important - rather than ponder and discuss 50 different causes, start with the most straightforward hypothesis and test it. Use a process of elimination until you are certain you have reached the root cause of the problem.

Should preventative maintenance Down Time be allowed to penalize an OEE calculation?

If the preventative maintenance happens during Planned Production Time (rather than during Planned Shut Down) then it is interfering with production and should be counted. This does not mean that you should not do preventative maintenance at that time; it only means that over time, OEE will tell you whether this maintenance results in less overall Down Time, because of fewer "surprises", or more Down Time, because of the lost productivity while maintenance is done. This is an example of how OEE can give you solid answers to complex questions by attaching them to simple measurements.

Should reworked pieces be counted as Good or Reject in the OEE calculation?

In OEE, any part that doesn't come out right the first time is a Reject Piece. It's a production issue that needs to be addressed. Your strategy for improving original part quality vs. reworking reject parts is a separate management and financial issue.

How are reworked pieces counted in the OEE calculation?

Reworked pieces are typically looked at as a completely independent OEE process. It is taken into consideration that time and materials were used in the first pass run that resulted in a part or product that did not meet quality standards but where a part of some value exists. Where the original Ideal Cycle Time may have been 180 seconds, a rework Ideal Cycle Time may be only 26 seconds, for example.

If a process can run during scheduled breaks, then is that break time factored into equipment availability?

A personnel break is not necessarily the same as a process break. If the process is expected to be running (i.e. producing anything at all) then regardless of whoever is or isn't eating lunch, OEE considers the time to be part of planned production. On the other hand, in a multiple operator non-automated process, a different Ideal Run Rate may be applicable during the break.

How does OEE relate to the management function of "enabling" operators?

The true value of OEE is in helping you and your operators make systematic improvements. Therefore everything you do with OEE, including the visual display of OEE data, should be designed with operators in mind and stated in whatever terms they most easily understand.

Click here for the OEE Glossary of Terms
Download the Fast Guide to OEE
Learn how Vorne can assist with your OEE initiative

back to top