Monthly Archives: July 2013

OEE sets the standard for improvement metrics

Which way to improve? OEE gives directions!One of the best ways of getting people to cooperate is to give them a common improvement goal. And OEE is the best example of what a good metric for improvement should look like.
The best example of cooperational excellence I’ve encountered so far is production teams working together to improve the performance of their production line. You could argue that the fact that a team of operators work on operating the same piece of equipment imposes the need to cooperate. The effect of improving together is much stronger, to my experience. And OEE – Overall Equipment Effectiveness – is, to my opinion, the best gauge you can give a production team for this purpose.

The theory of OEE
In short, OEE is defined as the time a production line is running effective (i.e. producing to quality standards at the rate it was designed for) as percentage of the total time that is available within a certain period. For example, for a a printing line producing printed forms OEE is calculated as follows:
Total production last week: 500 boxes of forms
Total time that would be required for production of 500 boxes, when the machine would have run at the speed it was designed for, without stops: 24 hours (= effective machine time).
Total time in the last week (opening hours of the factory): 120 hours (= total time)
OEE = effective machine time / total time = 24 / 120 = 20 %
But there is more.
OEE also shows you where the unproductive time went: the 6 big losses (of which there are actually 8)
Frist, there are2 Planning losses:
– Scheduled maintenance
– No orders or no materials available
Secondly, there are2 Availability losses:
– Set-up time
– Technical break downs of the machine line
Then there are2 Speed losses:
– Machine is running at lower production rate that it was designed for
– Short stops (paper jams, etc.)
And finally, there are 2 Quality losses:
– Machine is producing products that aren’t meeting the specifications (rejects)
– Products produced at the beginning of a run, while the machine settings are adjusted

What’s so great about OEE
One great thing is that all time is accounted for. It is impossible to manipulate the system. E.g., if you don’t register a technical break down, the time loss will end up in your speed loss.
Another great thing is that calculating improvement potential is easy (but not relevant for this story today).
But the very best thing is that a standard improvement approach is available for each type of loss. This helps the production team enormously in their improvement efforts. At any moment, they can start a kaizenning (improving) on the biggest of the 6 (i.e. 8) big losses, and by choosing the loss they want to address, they automatically know how to improve it.
For instance:
For set-up time reduction, there is the SMED method (Single Minute Exchange of Die): videotaping the set-up and eliminating all the muda out of the time that the machine doesn’t run.
Technical break downs can be prevented by introducing TPM (Total Productive Manufacturing)
The machine speed can be improved by – and this is a fun thing to do – step-wise increasing the speed until something goes wrong, watching closely what it is that goes wrong, and then finding a solution for what goes wrong.
And so on, for all the losses.


OEE sets the standard for all measurement systems
When you introduce a system of metrics, intended to help improving your team’s performance, OEE sets the standard. See if you can create a closed balance, where registration mistakes don’t effect the overall performance metric. And, more difficult but also even more valuable, design a standard improvement approach for each of your non-productive categories.
In my time as a Lean Manufacturing consultant, I’ve introduced OEE within numerous production companies, and it still is the standard for production lines. You don’t need a sophisticated manufacturing execution system (MES) for it (unless you’ve already got one, of course). When you know OEE and your production line, Excel and time registration forms are quite sufficient.
Introducing a similarly good metrics system in for instance service environments, that’s always an interesting puzzle!