In Part 1, we talked about the “why” of quick change. Now, let’s attend to the “how”.
I usually tell clients that, assuming changeover time hasn’t been attended to at all, they can expect about a 50% reduction pretty quickly, another 50% reduction with a bit more work and investment, and another 25% to 30% in the long run with continual improvement. So, if a changeover takes two hours initially, they can expect to reduce that to about 20 minutes or less. And I think that’s conservative. How can I be confident of such results? Because, as I reported in my previous post, my experience is that most organizations that haven’t attended to changeover time are pretty bad at it. It’s not so much that they’re bad at actually carrying out setups and changeovers (though, that may be the case sometimes). It’s more that they’re bad at analyzing and improving changeover times. Or even seeing the need to. But I went over all that in my previous post, so let’s get on to how to get those 50% reductions.
Before I start, though, I want to say a word or two about what we’re trying to accomplish here. I know…I use the term “quick change” and refer frequently to reduction in setup and changeover time. But our goal isn’t going to be improving speed so much as it’s going to be increasing consistency. In talking about why I seem to changing my focus from speed to consistency (and I’m not, really…we’re going to get speed by improving consistency), let’s step back and discuss the impact of changeovers (in particular) on scheduling, in particular.
And, by way of doing that, let’s look at an example from one of my past clients. This client had very complex changeovers of an entire roll forming line that sometimes took several shifts…when they didn’t take several days. When I asked the setup techs about changeover times, they reported wide ranges. A particular changeover might be completed in a couple of shifts one time but might take four or five days the next. As you might imagine, it was very difficult to stick to any kind of rational production schedule under these circumstances. They would have been better off if they completed changeovers in two days every time than they would completing some changeovers in two shifts while others took nearly a week.
So, OK…we’re going to get faster by focusing on getting more consistent. In other words, standardization is going to be the key to our early success. We’re not going to worry (overly) about how quickly a setup or changeover gets completed, rather we’re going to be fussbudgets about how smoothly our changeover goes and whether all the steps are readily repeatable.
I recommend two early steps:
1.) Have setup techs record time of last good part and time of first good part. Also have them make notes as to any problems or issues they ran into.
2.) Have someone just sit and watch a few changeovers (or portions of them if, like my former client, changeover time is measured by shifts). The observer won’t need a stop watch. A regular wristwatch will be handy to track some of what the observer sees; she might want to take note as to how long it took the setup tech to go back to the tool crib and find the right components, for example. A notebook will be helpful. You’re not trying to determine where every tenth of a second goes…you’re just trying to get a general idea of actions that take place that don’t add value.
In my next post, we’ll talk about what to do with your findings.