All you want to know about
AVAILABILITY Problems in OEE
10 FAQs about Availability Problems
1. Why is it called "Availability"?
Q: Why is it called “Availability”?
Daniel Högfeldt • By availability it is meant that the equipment should be available when it is planned to be. A customer does not care if the equipment is available at times it is not needed to produce their products.
What is tried to measure is the equipments reliability to be able to run according to the customer needs, not if the machine is able to run during all working hours during the shifts. Many companies try to always have a machine in work because they perhaps want to fulfill the demands on income from invested capital. This may however lead to a faster production than the takt time, therefore overproduction, the most significant source of waste. (Learning to See)
2. What are the categories of time within the availability of OEE?
Q: What categories (types) of time should we define within the availability of OEE (and OOE, TEEP)?
To calculate the ‘availability‘ of the equipment, time is categorized in the following types:
- Is the equipment running? Something is coming out?
- If not:
- Was it broken down, was there a technical failure to the equipment?
- Was there an organizational reason why it did not run; it was technically ok, but had to wait anyway?
- Was there a logistical reason why it stopped?
→ Line Restraint
- Was it taken out of service during normal shift operation?
- Was it not scheduled to run at all? No shift planned?
→ Not Scheduled
- Was it broken down, was there a technical failure to the equipment?
The goal is: To categorize events according to its major-loss type.
A machine is running
when there is output,
regardless of the quantity or quality
3. How many different 'Failures' should we define?
Q: You want us to track a maximum of 10 to 15 failure categories, but I want to define 85. How else can I ever know what part is breaking?
Arno Koch • At first glance that seems to make sense. But OEE is not a breakdown registration system, OEE is a loss detection system.
Let me explain the consequences: Imagine after 3 months registering OEE data, it shows the main loss is in availability. So now you want to know what you need to do to get your availability up. So you take one of the seven tools and draw a pareto diagram of all your time-events. Lets assume each of the 85 breakdown items occurred at least once. What you will see, is a pareto with an immense long tail, not giving a clear clue where the main losses are located.
Register process steps instead of components
In the other approach where you would have registered failures on, lets say, 5 process parts of the equipment (like ‘incoming conveyer, pre-heater, moulder, compressor, outgoing conveyer’) it would reveal what part of the equipment is restraining the process most. Then you start temporarily to focus on that part. A simple registration card (maybe even showing the 85 breakdown items) for some weeks will give detailed insight what’s going on.
Sometimes it is a matter of some simple maintenance, sometimes a Small Group Activity can solve the problem for once and forever. By this circle of focused improvement the equipment will become better and better.
This example also shows another disadvantage of the ‘breakdown registration approach’; if you handle each breakdown in the best way, thus eliminating it forever (either by reengineering or taking it into PM), it will not reoccur, but other breakdowns may start to occur. So after a while the breakdown registration system will not reflect the reality anymore. In the other approach this is less likely to happen.
4. What is: 'Loading Time' - 'Shift time' - 'Potential Production Time'?
Q: I am a bit confused about Loading Time, Shift time, Potential production time. How do they relate? Or are they all the same?
Arno Koch • ‘Potential Production Time‘ is the ‘general container’ that is to be determined.
Normally, potentially, the machine could run the whole shift.
And OEE -being a shopfloor tool- it is normally calculated per shift. Later, shifts can be aggregated.
So potentially, within OEE the shift-duration is the 100% limit for the availability.
Now in some specific situations, it can be decided to un-schedule a machine, e.g. in cases where there is no demand.
In such cases, the shift-time is reduced by the un-scheduled time.
‘Loading time’ is normally defined as the time the machine is supposed to be scheduled; so normally it equals the ‘shift time’, sometimes reduced by the un-scheduled time.
If we follow that path, maintenance, setups, and many other events are planned and would not be included in the OEE.
Of course in that way, OEE loses its value. Why try to measure it if you exclude all kind of losses…
Now an interesting question rises when we consider the OEE value during a shift. What now is the duration against which the OEE is being calculated?
Please see also:
5. How to calculate Availability (and OEE) when a shift is not completed yet?
Q: What is the availability (or OEE) of a shift when the shift is not completed yet?
In the book ‘OEE for the production team‘, you mention the shift time as the basis for the calculation.
I entered some data in our OEE software (OEE Toolkit v6). It looks like the availability is calculated based on the time which was actually registered (see Pic below).
So here it takes the 15 minutes the shift is on its way as 100%. Now 26,7% of that time the machine was running.
But in the book, you mention the shift time as the basis for the calculation.
So I expected the total shift-duration (510 min) would be the reference value for the availability.
In that case, 4 min out of 510 would be the availability (0,8%)
When is the full scheduled shift time taken into account?
Arno Koch • The question you raise is: How is OEE calculated during a shift?
If the OEE is to be calculated for a shift, (let’s say a shift with a duration of 510 minutes), what then is the OEE after 120 minutes?
Here we can follow 2 strategies:
1. Take the 120 minutes as 100% and calculate OEE
2. Take the 510 minutes of the shift as 100% and calculate OEE
In the situation 1, typically OEE jumps up when running and drops down again when idling.
In situation 2, OEE will start low and gradually evolves to its final value.
I prefer situation 2 since it is more intuitive to the operators. After a while they know what is a normal and what is an abnormal behavior of the number.
If well presented in a graphical manner, the team sees the ‘still to go’ part of time melt away, and the OEE pattern appears.
It is easier to steer upon than a number that has no reference point in time, other than ‘now’ and that jumps up and down.
6. Should 'Breaks' be in- or excluded in the OEE?
Q: Why do you include breaks in the OEE? We have a legal right to have a break!
Arno Koch • OEE measures the machine, not the operator. During the breaks, the machine waits for something to be done before it can start generating money again; to the machine a break is a loss. The machine does not need these breaks, even if the operator does…
The machine does not requires breaks!
Why stopping a machine during the breaks?
Your approach assumes the machine can not run while you are having your coffee or lunch. The reason to include it in the OEE is to make the production team aware of this potential loss. Is there really no way to let the machine run 15 or 30 minutes without operator interference? Would it be possible to have an other operator at the machine? Could an other operator watch your machine while you go away? Could you load it and just leave? Why not??
Show breaks in TEEP
Another approach to make this loss visible when taken out of the OEE, is to show it in the TEEP (TEEP measures the effectiveness including ALL time, 24/7).
TEEP is absolutely to be observed simultaneously with OEE, but one should carefully consider what behavior from management and from the production team is required and desired.
Use the creativity of your teams!
My experience is that operators can be extremely creative and more than once found solutions to keep it running. This behavior was initiated because they saw the loss and thus the improvement potential in their OEE! Achieving 6% improvement in effectiveness at no cost and little effort was worthwhile thinking about for them without a manager who forced them to change their routine…
So this is a matter of perspective or believe: who could and should respond to such losses.
In my perspective, the operator is the psychological owner of the equipment: He is the one who is the professional in the art of repeating the same value creation over and over again with his machine. He therefore should also be able to request the right actions he needs (either from himselves or from others that are there to support the value-creation) to be able to do so.
7. Should 'Preventive Maintenance' be in- or excluded in the OEE?
Q: Why do you tell us to keep Preventive Maintenance within the OEE? I, the operator, do not have any influence on preventive maintenance.
Arno Koch • You, the operator, are part of a production team. Together with your maintenance and engineering colleagues you are responsible for the effectiveness of the equipment. If you, the operator, can prove with facts and figures that you are suffering severe losses due to repeated breakdowns and too little (preventive) maintenance, or due to too frequent PM, it serves a mutual interest if you bring this up in your production team meeting!
Here too: If PM reduces your effectiveness: stop it, otherwise: do it!
8. Should 'Cleaning' and 'Maintenance' go into the OEE?
Q: We feel cleaning and maintenance should not be included in OEE. This is necessary to keep the machine running well!
Arno Koch • Exactly! So Maintaining a machine is not meant to reduce its effectiveness… No, we clean and maintain to RAISE its effectiveness. By taking this time out of the OEE, we will never see if the effort we spent to clean and maintain is bringing us a higher effectiveness at the bottom line!
The same of course applies to setup’s: It is crucial to be able to perform an effective setup in order to have the right output as the customer demands it. Hiding it form your OEE would take away the focus from this important issue!
Read more at:
9. What to register when we start 'Cleaning' while we 'Wait for Technician'?
Q: If we have no raw material, or we are waiting for a technician, we start to clean. What do we register?
Arno Koch • The reason why the equipment stopped running is not the cleaning but the ‘waiting for technician’.
Always register the true reason why the machine is stopped, not how you spent that time.
10. How do we treat a coil change at a slitting machine?
Q: We are slitting coils into steel bands. How should the time needed to change coils be integrated into the cycle time of the slitting machine?
Situation: A large coil of steel-foil is placed in the slitting machine. The steel-foil is un-winded and slitted in several steel bands. At the end of the coil a new coil must be placed in the machine and the process goes on.
Arno Koch • The goal of OEE is to visualize ALL losses. When the machine is not slitting although there would be a demand, this is a loss. Even if (you think) you cannot avoid it. This loss can be visualized in two manners:
- As waiting-time in the availability or
- as a performance loss (it would be considered to be a minor stop).
Make losses visible!
When the changeover time would be settled in a (reduced) cycle time, the loss would not become visible, and probably changeover time reduction will hardly become a subject of improvement activities.
As a rule of thumb, usually a waiting time larger than 1 to 5 minutes is to be considered as an availability loss. Since a coil-change normally takes longer than a minute, I would suggest to register this as ‘coil changeover time’ and visualize it in the category ‘waiting time’. (Meaning: The machine is waiting until the coil is being changed). In that way it is possible to see how many times we change a coil and how many minutes of potential production time is being lost. By applying SMED or eventually modifying the equipment, this loss might be minimized.
Cycle time slitting machine – Conclusion:
The cycle time of the slitting machine should be base on slitting and slitting alone. This is it’s sole purpose!
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.