All you want to know about
EQUIPMENT EFFECTIVENESS KPI’s
6 FAQ’s around other Effectiveness KPI’s
1. Why are there several (OEE) Effectiveness KPI's?
“(Why) do we need different OEE KPI?” When the OEE Industry Standard team started, there were many similar discussions and argumentations heard.
A basic statement was: OEE measures the effectiveness of a machine, the equipment.
Is OEE a KPI for Operations Effectiveness?
If well defined, this of course can tell us a lot about how effective the organization uses this equipment. But also about the surrounding equipment and even the supply-chain. However, OEE is not the KPI for “operations effectiveness” or “Asset Utilization”, although it’s loss-structure is an essential part of such KPI’s.
The OEE KPI is a shopfloor tool
The reason for this decision is the principal choice that OEE is a shopfloor tool, a metric that can be influenced by the shopfloor team. Of course, now there are some ‘grey zones’. How far can the span of control of a shopfloor team stretch? And what is the definition of a ‘shopfloor team’?
One dataset serves all!
Nevertheless, using the same (OEE) data, yet stretching the time-frame more or less, can result in a different ‘scope’ that may be interesting to a different audience. Particularly to the operations- and upper management.
To clarify what we are looking at, what scope is used by what audience, we defined some different scopes that are commonly extracted from OEE data.
You will see several different formats of the same loss-structure data set.
I agree this is all a matter of nomenclature, but this diagram shows that we are talking about the same loss-data structure. It is just a matter of reshuffling it in different ratios, for the purpose of gaining focus to different audiences.
OEE KPI ‘s to create a unified language
Important is to create a clear ‘language’ and to use each tool for its most suitable purpose. We want to prevent the ‘abuse’ of tools for different purposes just because it is there and could be used. (You will understand I hate to see Bahco wrenches and pliers in the factory for the same reason.)
Although slightly outside the scope of the OEE Industry Standard, the different ‘scopes’ are part of the standard. However, they were never subject to serious discussions and reviews.
It seems for the new 2020 standard this would be a fine moment to do so. I would like to invite you to review the currently defined scopes and comment them:
- Are they clearly enough defined?
- Do they fulfill your needs?
- Are there already other standards that could be included here?
To do so a dedicated (open) group is started to discuss the standard, http://www.linkedin.com/groups/OEE-Industry-Standard-4246823
I hope together we can further improve the standard.
2. OEE - OOE - TEEP explained
Q: What is the difference between the OEE, OOE and TEEP measurements?
Overall Equipment Effectiveness measures a machine’s availability, performance and quality in order to visually display all the losses that occur on this machine. So how do OOE and TEEP measurements differ from this?
Arno Koch • The three measures are all visualising effectiveness losses.
The only difference between OEE, OOE and TEEP is the ‘maximum time’ that is used in each calculation.
In other words, the only changing variable between these three calculations is the maximum time that is available for a machine to run. They all take availability, performance and quality into account.
Total Effective Equipment Performance Considers maximum time to be All Available Time – that is 24 hours, 365 days a year.
Overall Operations Effectiveness takes unscheduled time into account, looking at Total Operations Time as a maximum.
Overall Equipment Effectiveness looks at the Potential Production time as a maximum, without calculating time that has been unscheduled.
Introduction: what is behind those KPI’s ?
In the ideal world, manufacturing equipment would be adding value all the time, at theoretical maximum speed, in precisely the desired quality.
So, if we want to know why our equipment is not ideal, there are 4 questions to be answered:
- Was the machine scheduled to run or not? The time it is scheduled to perform is also referred to as shift-time or loading time.
- When the machine was scheduled to produce, for what reasons was it not running? What was stopping the conversion process?
- When it was running, how come it was not running at theoretical maximum speed? What was reducing its speed?
- In the produced output; what quality defects occurred?
‘Scheduled’ versus ‘not scheduled’: Loading Rate
Please note: ‘Scheduled Time’ is also referred to as ‘Loading Time’
So first we determine: How much of the total time was the machine scheduled to run?
This is the Loading Rate:
Loading rate = Scheduled Time / Calendar Time
Please note: Breaks and other waiting is INCLUDED in the loading-time:
It is part of the operation; theoretically the machine could run!
Once the machine is scheduled, it will usually be operated in a certain shift regime. However sometimes it will be unscheduled again. Regular shift time can be unscheduled for several reasons; ie there is no demand, R&D needs to runs some tests etc. Those reasons are beyond the scope of the production-team.
Please do not confuse un-scheduled with not-scheduled or unplanned!
‘Planned’ versus ‘Unplanned’
Preventive maintenance usually is planned. Corrective Maintenance is usually unplanned. Change-overs may be planned, so are breaks.
Of course we rather see planned waiting times than unplanned ones. Yet to the OEE it does not make a difference. The machine waits, is not effective. For purposes of analyzing- or visualizing the losses it may be interesting to distinguish ‘planned’ and ‘unplanned’ stops, yet planned or unplanned does NOT results in a different OEE, OOE or TEEP
Running or not: Availability
A machine is scheduled (‘Loaded’) in order to produce. When the machine is scheduled, we measure: “Was it having output, regardless the speed and the quality?”.
The relation between the time the machine was having output and the time it could have produced output is called
Availability rate: The percentage of time the machine was available, was able to produce output.
Whenever the machine is running, thus really produces output, we determine how much, or how fast, the output was produced.
The ratio between the theoretical quantity of output while running, versus what actually came out, is the Performance rate.
From everything that came out, it is determined whether it was meeting the quality as expected. Was the output within the specification?
The ratio between the total actually produced output and the good output is called the Quality rate.
There are several KPI’s possible around those three rates. The main difference is the time-frame that is been considered while looking at availability.
In the end they all those KPI’s want to know: How much good product came out versus how much could it have been produced? And here is the crux: it all depends on what time frame is considered.
If we look at a years’ production; is this now based on 365 days of 24 hrs? Or is it only the time of all the shifts we would have scheduled? Or only the actually scheduled shifts? Or do we only consider the time we planned the machine to run within the scheduled shifts?
The KPI cascade
Let’s break it down to the core.
Total Effective Equipment Performance (TEEP)
If we want to know the real utilization, the real effectiveness of our equipment, we look at the amount of good product produced versus what could have been produced if there would have been no restraint at all.
This is what TEEP does. It takes 24hrs of 365 days a year as potential production time.
A well defined TEEP (and this also applies to OOE and OEE) takes the theoretical output as a maximum for performance and OTIF (so within specification AND on time) as a quality limit.
TEEP calculates availability as a ratio between the TOTAL calendar-time during the measurement versus the time the machine was actually having output in that period.
So in a 1-shift operation of 40 Hrs during an 168 week, the availability can be theoretically no more than 40/168= 23% (assuming OEE=100%). TEEP now shows clearly that the equipment is sleeping at least 77% of its life time due to the fact it is not scheduled to produce. In this example TEEP = Loading Rate.
So one could also say:
TEEP = Loading Rate x Availability Rate x Performance Rate x Quality rate …or
TEEP = Loading Rate x OEE
Overall Operations Effectiveness (OOE)
For those working in the daily operation, TEEP uses a time-frame that is ‘too wide’. The operations management is interested in the effectiveness of the equipment during operations time. So the first filter to apply is to take out all the time that the equipment was not scheduled anyway.
In other words:
OOE uses all regular shift time as the time-frame.
Overall Equipment Effectiveness (OEE)
This is the KPI for the shopfloor-team.
OEE relates the actual good output versus the ultimate potential good-output while the machine was scheduled to run. (So during the actually scheduled shift time).
What if we have a line that runs one shift (8 hrs) and there is a production plan is for that time; what happens with the other two shifts? In theory the line is available to run 3 shifts, but its been used for just one shift.
- There is ONE shift scheduled: The OEE will be measured over that shift. Non scheduled time (two shift that day, plus the missing shifts in the weekend) are not included.
- There is no unscheduled time (time taken out of the scheduled time) so the OOE equals the OEE.
- The TEEP will be calculated over the TOTAL time, so the scheduled plus the not scheduled shift: it will be fundamentally lower.
Q: How does OEE compare to ‘TEEP’ and ‘Utilization’?
Daniel Högfeldt • It is important to measure utilization at the same time as measuring OEE. Here are a few different ways of doing that:
Asset Utilization is the percentage of the actual operating time and the total available time (calendar time, i.e. 24 hours per day).
TEEP (Total Effective Equipment Performance)
TEEP (Total Effective Equipment Performance) is the percentage of calendar time the equipment is running at speed and making good product. It is the ratio between the total available time (calendar time) and the actual effective production time. That is the same as the theoretical maximum amount of good products versus the actual amount of good products made during that time. It can be calculated by multiplying the Asset utilization, Performance efficiency and the Quality rate together.
Capacity Utilization is the ratio between the total available time and the loading time. It shows the hidden operations potential. For example it would show the difference between 2 and 3 shifts.
3. Measuring effectiveness of machines that are not always scheduled
Q: We run a 3 shift operation of 120 hrs a week, but some of our machines are not 100% planned to run due to the product mix we have.
How should we calculate the OEE?
During some product-mixes we do not use a certain machine because, for example, line A can run products in sizes 10mm to 50mm and line B can run products from 45mm to 100mm.
What would be the OEE in the following example?
- Machine A is planned to run only 80 hrs of it’s possible 120 hrs per week
- It has 20 hrs breakdown
- Assume the Performance and Quality 100%
Reading the site I expect it to be 20/80 but I’m getting disagreements at work.
Arno Koch • OEE is an instrument to measure equipment effectiveness from the perspective of the shopfloor team. So the shopfloor team measures OEE from the machine it is working on.
The machine is basically scheduled to run 3 shifts, so one would expect to have an OEE measured in these 3 shifts.
But: due to market-constraints (for product that can be produced on a certain machine) a machine is taken out of the shift schedule; it is being un-scheduled.
In your example it is un-scheduled for 40 hrs; there is no team at the machine, the machine is not running. So there is no OEE.
The OEE’s availability would now be:
Running time (60) / (Scheduled time (120) – Unscheduled time (40) ) = 75%
BUT…. The machine COULD have been running IF there were a different product mix or no market constraint! So this is a matter for the operations management to see and act upon.
That’s why –beside OEE– we also look at the OOE, the Overall Operations Effectiveness.
In the OOE we do not care about unscheduled, planned or not, we consider the total available shift time of the regular shift-schedule as 100%.
So your OOE’s availability would be:
Running time (60) / (Planned time (80) + Unscheduled time (40) ) = 50%
Or in other words:
Running time (60) / Total Shift schedule (120) = 50%
The owners of the facility could even extent this way of thinking and ask you: “Why is the machine only scheduled for 120 hrs? There are 168 hrs in a week; the machine could be operated in a 5 shift operation!”
That’s why there is TEEP, which takes 24/7 as 100%
So now your example would bring how much TEEP availability?
Running time (60) / (Planned time (80) + Unscheduled time (40) + not scheduled (48)) = 35,7%
Or in other words:
Running time (60) / total time (168) = 35,7%
Where is the difference?
OEE, OOE and TEEP are being calculated based on the same numbers. They only take a different reference value for “maximum time” and thus they will show a different loss-cascade.
The shopfloor-team will focus on the part they can and should influence; the time they are operating a machine that has demand. Operations management and higher management should support this ánd do their part also.
4. How do I best measure the performance of our machines?
Q: I have been asked by my boss to find a way for evaluating the performance of the Machines in our factory and I found the implementation of OEE is the best way. Now I have some questions about how to define OEE, I have been following this website which helps me a lot.
Arno Koch • It is good to hear this site helps you and I am happy to see you refer to your Machines as with a capital M 😉
Your question raises some new questions for me…
- What is it your boss is really looking for? Is he deeply interested in the performance of his machines? Does he want to compare them, and if so why?
- What will be the role of the shop floor team when the performance of the machines has become clear to your boss? What will your boss do with the numbers you are going to establish?
- Is your boss trying to find a way to compare the machines?
5. What is EFFICIENCY, EFFECTIVENESS and PRODUCTIVITY?
Q: What is the relationship between efficiency, effectiveness and productivity and what is the path that can be followed to bring about ‘improvement’?
Efficiency is determined by the amount of time, money, and energy – i.e. resources – that are necessary to obtain certain results. In order to meet our daily production quota, we commit a specific machine that uses up energy, make operators and maintenance personnel available, and provide raw materials. For example, if we are able to meet our daily production with less energy and fewer operators, we have operated more efficiently.
Effectiveness is determined by comparing what a process or installation can produce with what they actually produce; therefore, effectiveness does not tell anything about the efficiency – the amount of resources that have to be committed to obtain that output. If we are successful in manufacturing more good product in the same time period, effectiveness will increase. A valuable discussion could be whether ‘good product’ should be seen as ‘Good product with customer demand’ to prevent over-production.
Productivity is determined by looking at the production obtained (effectiveness) versus the invested effort in order to achieve the result (efficiency); in other words, if we can achieve more with less effort, productivity increases.
Goldrath (‘The Goal’) defines productivity as: ‘the extent in which a company generates money’. The goal of a production company is therefore not to reduce expenses but to generate as much money as possible!
How much productivity improvement is possible?
Most striking when looking at the traditional approach to improvement, is that focus is often exclusively on efficiency; the famous cheese slicer continuous to slice production further and further.
How much room for improvement is still left at the input (efficiency) side? 10%? 20%? And does it still make sense to try to reduce another operator or engineer, or to put pressure on the buyers to negotiate even more competitive prices?
As is often the case, that question can not directly be answered. If the supplier can give us a better price because we help him with managing his production process better or – as we see in the automotive-industry – we force the supplier to go through basic improvement processes, such as Lean Manufacturing or TPM, not only the price will decrease, but the quality and the reliability in delivery will increase as well. That is good news for both parties.
However, by managing solely by keeping the cost price down, you run a large risk of saving pennies per product but losing many euros or dollars due to stoppages, quality losses, etc.; in other words, ‘penny wise, pound foolish’. Many production teams can give you striking examples of that.
We strangely look less often at the output side – the effectiveness – of the equipment. Apparently, the output is more or less considered to be ‘as it is’. However, every line manager knows that the installation will spontaneously start to run better simply by standing beside it and giving it attention. When you check the logbooks, there are days that, on occasion, the installation produced spectacular amounts of good output.
It happened to go well that day…
Ask the team how that happened and you will hear a precise run-down of all elements that went right that day. The raw materials arrived on time and were of the correct quality, the installation kept on running and was set correctly, the right people were present, it was not too warm, etc., etc. This is often regarded as a fluke and nobody is wondering how you could create a similar situation a second time. That is strange actually, for if it can happen once, why should it not be possible to happen again. And if it can happen a second time, why not always? Usually, a whole series of ”Yes, but’s” will follow…
Suppose you would write down those “Yes, but’s” and turn them into a list of action items. What would that give us? To be able to answer that question, we will have to dive a little deeper into the world of Effectiveness.
Our machines run non-stop!
What determines the effectiveness of an installation? First of all we must address the question of whether it does or does not run. Roughly there can be three reasons why an installation is not running:
- The installation quit; it broke down.
- The installation could be running technically speaking, but is waiting for something; materials, an operator, filling, to be set-up, etc.
- The installation could be running, but is not planned in because there is more capacity than demand.
Of course, the ideal machine would never break down and would never have to wait for anything; therefore, it would be running all the time as long as there is demand for the product.
Our machines run at top speed!
Subsequently, the effectiveness is determined by the speed at which the installation is running. This is always a tricky topic, for what is the maximum speed? The speed at which it is at the verge of breaking down? Or the speed at which the quality of the output reaches the bottom limit of its spec? This will guarantee a lively discussion. It is useful to see that it is often simply unknown what the maximum speed is, while the maximum speed that people come up with is usually based on various assumptions (which, in turn, you could turn into an interesting list of action points!). An example of such assumption is: “If I ran the machine any faster and the material got jammed, we would suffer major damage”. Why does it get jammed? Is that always the case? When not? What must happen to prevent it from jamming again? Why does damage occur when it gets jammed? What do you have to do in order to…. etc.
We have ‘zero defects’!
If the actual speed is determined versus the theoretical speed, the next effectiveness determining factor can be looked at: Does the realized output meet the set quality standards? It can be quite an eye opener, if you ask ten different people on the shop floor to indicate very clearly when a product does or does not meet the specs, you will receive ten different answers. It becomes even worse when it turns out that the one who produces the product, the operator, cannot determine, or cannot unequivocally determine this. Also here lie many opportunities to solve all “Yes but’s” and to ensure that the person who makes the product is also able to determine whether he is manufacturing a good product, so that he can keep the quality within pre-set specifications.
Are we ‘ideal’?
Thus, if there is demand, an ideal and effective machine is always running at maximum speed without producing any out-of-spec products. We can assume this to be 100% effective. We know that 100% effectiveness is impossible over a longer period of time; after all, installations must sometimes be maintained and converted. The guideline is that 85% is a realistic “World Class” value for “traditional” machines. That implies that the installation, for example, produces 99% of the products “First Time Right within Specs”, operates at a speed of 95% of the theoretical maximum speed, and is actually running 90% of the operating time (99% quality x 95% speed x 90% running time = 85% effectiveness).
In a three-shift system it means that the installation runs for 90% x 24 hrs = 21:36 hrs at 95% speed with 99% quality. Consequently, there will be 03:24 hrs available for maintenance, conversion, and other possible waiting times. Incidentally, the 85% mentioned is a rather conservative figure; nowadays we see in the automotive-industry equipment that runs over 90%.
The analysis of hundreds of installations for various processes shows that, as rule of thumb, an average installation in an average (non-TPM) company runs at an effectiveness rate between 35 and 45%. Of course, there are always cases that stick out; for example, values in the pharmaceutical industry may lie considerably lower and there are also cases that show considerably higher values.
If it turns out that an installation has an effectiveness of 40%, while people always thought that there were limited options left for potential improvements, it is extremely good news: this means that twice as much good product can be manufactured (your effectiveness rate would be 80%!) at the present cost level. Or, you manufacture the same product with one shift instead of two.
Yes, but then the costs will rise!
It is often assumed that achieving such improvements will necessitate an enormous increase in costs for, for instance, maintenance. That is sometimes partly true, for example, when it concerns overdue maintenance and you are then actually paying off a loan, because a fundamental design flaw has to be solved (and, therefore, you can also see this as paying off a postponed cost item). However, by activating the knowledge that is present on the shop floor in the right way, 80% of the improvements can often be implemented without any capital expenditures and at minimal costs.
It is not so hard to imagine that an installation, which halts on a regular basis for various reasons, or whose process is not stable enough to operate at high speed without any losses in quality, automatically requires more resources at the input side as well! Reversely, it may be that lowering the efficiency (for instance, by spending a little bit more money and time on preventive maintenance) will bring about a strong increase in effectiveness, which – bottom line – creates a higher net productivity. Such considerations can only be made if, in addition to efficiency, particular attention is paid to losses in effectiveness as well.
World Class Manufacturing
In all cases, it is necessary to take decisions concerning actions leading to improvement on the basis of facts and figures describing the entire productivity picture.
World Class Manufacturing does not accept any losses at all. That is what management must focus on and management must have the will to go further than mere window dressing and scratching the surface. Unfortunately, that is often even harder than just opening the wallet. Companies that do take this route, discover over and over again: There still lies a nearly unlimited potential for improvement for those who learn to see it and seize it!
6. (How) does OEE fit with Six Sigma?
Q: I was wondering if there are companies that have used the OEE information for Six Sigma projects and what were the results?
Robert Strout • Most industries are striving to eliminate waste. I have been using OEE software. Our company started Six Sigma and used the OEE information for the basis of it’s initial projects. We found that the information from our system was lacking because of the general areas of loss. The toolkit would allow a accurate definition of losses. I was wondering if other company has used the OEE information for 6 Sigma projects and what were the results?
Arno Koch • I worked for a company producing baby milk powder. They monitor their quality using 6 sigma principles. To stabilize their process they implemented process teams. Those process teams focus on several performance parameters; the OEE parameters are used to reduce the losses and get the manufacturing process stable. TPM is their framework to create stable machines.
They concluded that OEE is very useful to get shopfloor steering information there where the process can be improved: at the shopfloor!
Answer still not found?
To help you best, please be as specific as possible and please understand:
We do not want to reply to anonymous questions.