OEE as a performance KPI

How to evaluate performance across the departments of an industrial plant and use OEE as an incentive to find better ways to reduce or eliminate production losses.

Looking for an industry-independent "out-of-the-box" OEE calculation tool?

OEE software

The challenge of OEE

by Marc Leroux

I was talking to a customer, I’ll call him Fred, the other day and the topic of OEE (Overall Equipment Effectiveness) came up. We both acknowledged that OEE is now universally accepted as a significant Key Performance Indicator (KPI) for all manufacturers. Fred then raised two significant points.

“The first problem was that nobody liked using it because the number is too low. We’re used to judging performance in the range of 99%, and we’ll never get there with OEE. We’ve solved this, but it was a significant challenge initially.

The second problem is that our executive management discounts it because it is a manufacturing metric, not a business metric.” That launched a good discussion, so let’s take these points.
If we are used to judging performance in the range of 99%, we'll never get there with OEE. Therefore many people still don't like using this KPI because the OEE number is too low.

OEE calculation

Most people probably understand the components involved with, and the calculations behind OEE, but let’s recap. The OEE calculation is pretty simple in itself:


OEE = Availability x Performance x Quality

Availability, Performance and Quality describe the six big losses. Availability is reduced by equipment failure, setup and adjustment. Performance is reduced by idling, minor stops, and reduced speed. Quality is reduced by process defects and startup losses.

Because the numbers are multiplicative, if any one component is low, then the overall result is going to reflect that.

Availability

Traditionally availability was the responsibility of the maintenance department, and they were proud of their results.

“They used to hit uptime targets that were in the 98%-100% range, and the OEE calculation is putting Availability in the 70’s.”

In the OEE calculation, Availability starts at 100% less planned downtime which gives us our gross operating time. Planned downtime, or Planning Factor (Pf), includes planned maintenance activities and planned gaps in the production schedule, for example, weekends or shifts that are unmanned. The availability rate (A), or Net Operating Time, can be improved by the reduction/elimination of breakdowns together with reduction of set-up and adjustment losses.

Fred pointed out that initially they weren’t collecting the setup losses. “We had those built into our schedule as planned downtime.” I asked if that was really useful in driving improvements and he responded “No, not at all. There was no incentive to find better ways to reduce or eliminate setup or transition times. But if we didn’t include them in planned downtime the numbers looked pretty bad.”


The availability rate (A), or Net Operating Time, can be improved by the reduction/elimination of breakdowns together with reduction of set-up and adjustment losses.

Performance

This is based on subtracting losses from operations, such as running at a lower rate than a machine is specified for, idling time and minor stops. This gives us the Performance factor (P) and subtracting these losses from the Net Operating Time gives us the Valuable Production Time.

There was some lively discussion around performance. I asked what they were using for the production rates and Fred responded “Well, we started with the theoretical maximum rate, but quickly switched to our standard planning rate. I know that’s cheating. We implemented OEE to drive our operational excellence, so we should be setting targets around what we want to achieve.

My view is that I bought the machine to run at a certain capacity, and then we say that it is OK to run it at a slower speed. That doesn’t seem right. I tried arguing that we should use the design limits, or, at a minimum the best production rate we have ever achieved, but when we did that the numbers aren’t so good.”


"We implemented OEE to drive our operational excellence, so we should be setting targets around what we want to achieve. "

Quality

Quality (Q) represents losses from off quality production and as well as losses during startup/shutdown.

“We had a lot of fun with this one” Fred said. “As soon as some of the guys realized how we were calculating this they pushed some of their off quality stuff to the next processing station. That way their number improved, while the next station’s went way down. This was the classic case of having a KPI encourage poor behavior. We had to make some changes to ensure that losses were being recorded properly.

Then we got into the discussion on why startup losses would be included. The operations people argued that they had no control over losses during startup and they didn’t want to be penalized by them.”


“Now we are looking at things across department boundaries, and that’s just not comfortable for everyone. We have a measure that isn’t under the control of any one group.”

In the next part of the series - OEE as a business KPI - learn how to evaluate total productivity of an industrial plant and encourage executives to use OEE as a business measure that reveals opportunities and drives real improvements


OEE meaning

  • Contact us

    Submit your inquiry and we will contact you

    Contact us
Select region / language