I tweeted a link to an Automation World article on OEE in Packaging Machines and immediately started getting dissed by @RickBullotta and @simonfkj.
I’m well aware that OEE leaves a lot to be desired–but it all depends upon what your objective is. @simonfkj worries that there is no link to profitability. He’s right. But that whole thing is going to require a lot more software than most companies have. Heck, I think most are still using Microsoft Excel.
My problem with OEE (overall equipment effectiveness) is that many companies use it for comparison metrics and most do not apply it well. If they are going to use it as one, I hope, KPI, I would hope that they implement it well.
OEE is the product of three ratios: availability, quality and throughput. Even when I first heard about it in the early 90s (my customer, a Ford plant, had an OEE specialist of all things), I thought it was too vague. Then when I heard that people were using it to compare machine to machine, vendor to vendor, plant to plant, I thought “uh oh” this could lead to problems.
The thing is, each of the terms in the equation is subject to interpretation. For example, when is the machine available? Should you not count certain downtimes–say for scheduled maintenance? What is quality? Actually, even what is throughput? Unless management can define those terms and enforce the measures, then the OEE number is essentially meaningless. At best it could measure performance of one machine over time–and if I were plant manager, I’d even be suspicious of that.
But like I said, if people are going to use it, at least they should try to use it well. And even then, only as one metric among several.
In fact, I just finished interviewing an SVP of Schneider Electric North America on Lean and he said something that was cool. At the end of every line is a white board. If a line is not making numbers, the barriers to making the numbers must be determined. Anyone can go to the whiteboard and write down what they see as that barrier. If it is something that can be fixed immediately, it is. If not, then it is taken to the people who can make it happen.
When I was a rookie, I was taught that accounting was “ancient history.” What this person told me, in effect, is that even “real-time” information going into an MES is really “ancient history.” The white board is the actual real time. I like that–even if I am editor of Automation World.