OEE Rescue: OEE is composite and does not tell much per se

Overall Equipment Effectiveness (OEE) is probably the most widespread and well-known among KPIs in industry, which does not mean that everyone likes it. OEE rescue is a series of posts that aim to balance the love-hate comments about this KPI as well as debunking some myths and misconceptions.

In this post: OEE is composite and does not tell much per se

Yes, OEE is composite. OEE is expressed in a single dimensionless value. It’s a ratio, a multiplication of 3 other ratios (availability, performance and quality).

Not familiar with OEE? Follow this link

What I immediately liked when I discovered OEE is the fact that multiplying 3 fractions leaves the result smaller than the smallest fraction, meaning it is a very aggressive and challenging KPI.

Any worsening of one of the 3 constituent will amplify the worsening of OEE, which in turn should trigger quick countermeasures to stop the KPI to plunge.

I do not agree OEE per se doesn’t tell much. The original intent, I assume, was to provide management with a single value in order to get an instant feeling about how the overall performance stands, as well as a convenient benchmark when comparing machines, lines, workshops or factories. And it does the job well..

This advantage of being synthetic is also a drawback as it is necessary to “disassemble” OEE to its components in order to understand which of the availability, performance or quality is the evildoer.

But again, I see here an interesting “constraint” for management: the head of department will review OEEs and get a broad feeling about how well the various lines or cells of his/her realm are doing. When intrigued or alarmed by a poor OEE, he/she will turn to the supervisor or line leader to get more information.

The latter needs to know more precisely what’s going on as it is his/her responsibility to keep OEE at best. This required dialog is, from my point a view, a good way to have management commit to interact in both directions: top-down and bottom-up.

I suspect that the managers not liking OEE struggle to drive and maintain theirs on the expected level. Instead of looking how to boost their OEEs, they probably prefer criticizing the concept.

It is one thing to display a quality rate of 93%, a machine availability of 90% and a performance rate of 95%, which at first glance look good, and another thing to report a OEE of 79.5% which is exactly the same (0.795=0.93×0.90×0.95), except for the perception.

Yes, OEE is humbling.

By the way, there are other KPIs that are composite like the On-Time-In-Full (OTIF). When OTIF is bad, is it the On-Time or the In-Full part that hurts? You don’t know until you dig deeper into the details. Would you dare replying to your furious customer measuring your performance with OTIF that this KPI is composite and does not tell much per se?

View Christian HOHMANN's profile on LinkedIn

Advertisements

The very minimum to know about Overall Equipment Effectiveness (OEE)

Overall Equipment Effectiveness (OEE) is a KPI which reflects overall equipment or machine performance in a single number, expressed in %.

OEE compares the actual ok output to the total output achievable in perfect conditions. As perfect conditions are not likely to be permanently granted, OEE is a mere theoretical benchmark, nevertheless something operational staff should aim for.

Overall mean encompassing  three other key metrics: Availability, Performance and Quality.

  • Availability is the readiness of the machine / equipment to operate when required
  • Performance is the actual run rate compared to the nominal run rate
  • Quality is the number of good parts or quantity right first time compared to the global quantity (good and no good)

Everyone of these metrics is expressed in %, OEE = availability x performance x quality

As OEE is multiplying fractions, the result cannot be greater than the smallest value of Availability, Performance or Quality. That is why OEE is a severe KPI: if one of the intermediate KPI decreases, the OEE decreases faster.

OEE came with Total Productive Maintenance (TPM) and aimed primarily to maximize assets’ yield. OEE was designed to grasp the real performance and not being fooled by tracking assets’ utilization.

By utilization I mean the ratio machine time used / available time. A machine used for production 90% of the available time, seems good. But if this machine runs at 60% of its nominal speed and 15% of its output must be discarded, OEE will be 90% x 60% x 85% = 46 % which is not good at all, as more than half of its productive capacity is wasted.

Demonstrated capacity

The OEE performance is a snapshot of demonstrated capacity, e.g. what the machine / equipment / supplier can really achieve.

The capacity above demonstrated capacity up to nominal maximum capacity is wasted.

It is like a beaker (fig.) having a 100ml capacity but all the leaks limit the real capacity to 60%.

Changeovers are usually a major cause of machine downtime and account as “leaks”. Changeovers duration can be drastically reduced with SMED, thus fixing the leak and recovering some or the wasted capacity.

When breaking down all causes leading to waste of capacity, it soon becomes clear that maintenance issues account only for a (usually small) fraction. OEE therefore gained interest for itself, not only as a KPI within Total Productive Maintenance (TPM).


Bandeau_CH38View Christian HOHMANN's profile on LinkedIn

Overall Equipment Effectiveness (OEE) is a Goal Tree

Overall Equipment Effectiveness (OEE) is a well-known KPI in many industries for nearly forty years in western countries. It reflects overall performance in a single number, and is built upon or integrating three other metrics: Availability, Performance and Quality.

  • Availability is the readiness of the machine / equipment to operate when required
  • Performance is the actual run rate compared to the nominal run rate
  • Quality is the number of good parts or  quantity right first time compared to the global quantity (good and no good)

Everyone of these metrics is expressed in %, OEE = availability x performance x quality

As OEE is multiplying fractions, the result cannot be greater than the smallest value of Availability, Performance or Quality. That is why OEE is a severe KPI: if one of the intermediate KPI decreases, the OEE decreases faster.

If someone is in charge of improving OEE, he or she will “speak” a Goal Tree, and it goes like this: in order to achieve the highest value of OEE, we must have:

  • the machine / equipment steadily ready to operate
  • running continuously at nominal speed
  • producing only good parts

Now with this said, what to do next? Where are the leverage points?

OEE is great to give a snapshot of performance taking into account the three high-level must haves, but when it comes to action OEE must be broken down to find where and what to act on.

In Logical Thinking Process lingo, Availability, Performance and Quality are Critical Success Factors (CSFs), high outcome objectives that enable achieving the Goal: the highest OEE.

Critical Success Factors are very useful because they provide top management the minimal but sufficient dashboard to monitor progress towards the Goal.

These Critical Success Factors are dependent on underlying Necessary Conditions.

Availability for example depends on machine’s technical condition as well as on setup and material availability. It depends also on availability of trained and entitled workforce, work orders and possibly other documents.

OEE Goal Tree

Example of OEE Goal Tree

Performance will probably depend on the machine’s condition, itself depending on the maintenance policy, maintenance frequency, and so on. Performance is also depend on the proper use of the machine by workers, the type and quality of raw material, the type and condition of its tools.

The breakdown goes on this way, from each Critical Success Factor down to the least Necessary Condition, building the whole Goal Tree. In order to list all Necessary Conditions, the same phrase applies: “In order to achieve… We must…”

The list may go on, both horizontally and vertically, according to the business.

The more regulatory constrained the business, the more likely to have a horizontally-wide Tree, as those regulatory constraints will add many mandatory Necessary Conditions.

What I really like with Goal Trees is the provided guidance by the necessity-based logic, insuring a complete and robust Tree is built. On top of it, it helps discriminating the must-haves from nice-to-haves.

Therefore it’s easy to respond logically to the claim “the machine is too old to achieve good performance, get us a new one!”. When considering what can be done to improve OEE, the age of the machine does not appear as being the biggest influencer.

In fact, many examples show that properly tended old machines can still be performant assets.


Chris HOHMANNView Christian HOHMANN's profile on LinkedIn