I was at my box maker client conducting a quick change workshop yesterday.  We've already conducted a couple of 5S workshops and a previous Quick Change workshop in the same area; the response and energy of the participants (one of whom is about to retire) has been gratifying.

I've often conducted an observation of a changeover during these workshops. By that I mean, some of the participants carry out a changeover while others observe and take notes.  (I like video but in this case, it was difficult to get for a variety of reasons.  In any case, I like the observation method because it gets a number of people engaged.)

The only downside of such observation is that the folks carrying out the changeover invariably "hurry".  I can tell they're "hustling" a bit, trying to move a bit more quickly than normal.  I always ask on of the observers (who is usually someone who, themselves, is familiar with the changeover) if it looks as if the participant is moving faster than usual and the answer is generally, yes.

It turns out that this isn't much of a problem, though.  This "hurrying" usually brings the changeover time in within a few minutes either side of what they've told me in the classroom is a "typical" time for the changeover (assuming something didn't go wrong).  Yesterday, the "hurrying" led to a changeover of 45 minutes, exactly what they had stated a typical changeover might take.  The point is, trying to do a changeover faster doesn't work that well and that's when everything goes perfectly.  It gives me the chance to provide a learning moment: we make the changeovers quicker, not by moving more quickly, but by changing the way we do changeovers.

In yesterday's case, actually changing over the machine took about 15 minutes.  Tweaking the machine to get a good box took another 30 minutes.  In the long run, the team and maintenance is going to index the settings on the machine and have them included on the work order.  In the short run, after being told that they would be actually doing that particular changeover many times in the next few weeks, the team went back out, took some measurements and wrote them down on masking tape attached to the machine.  In other words, they'll be able to set the machine to the precise dimensions the first time through.  I'm guessing the process of setting the machine will take 20 to 25 minutes and (it's hoped) the tweaking will be reduced to 5 minutes or so.

Then the team will need to develop standard settings for all the changeovers done on that machine.  And that's the hard part.