I say: don't optimize your setups or change-overs !

on my blog about not reducing future forecast, I received many differing opinions and there were lively discussions going on. That is the kind of stuff that makes us think a little bit longer about our opinions and usually leads to a better solution... and then, of course, there are many different situations and each situation calls for a slightly different approach.

alright... here is yet another opinion of mine: "optimizing setups or change-over time in the production schedule adds to the waste of overproduction and is unnecessary!"

take a production line that has a capacity of 1000 hours in any given period. A forecast calls for the provision of certain quantities of products A, B C and D in that same period. If the production of those quantities can be done in less than 1000 hours, you have free available capacity. Typically, a costing person now says 'fill the lines' and you see that you can put more product on the line and try to utilize close to 100%.

Now, if you also optimize your change-over times, you will free up even more capacity and the production manager is busy to put even more product onto the line, which no one buys.

In my opinion, if you have left over capacity, you are better off just shutting down the lines rather than making product that gets stuck in the warehouse... now here is a radical idea: what if you don't shut down the lines but instead using the free capacity for more change-overs? what I actually suggest is that you cut down on the lot sizes, which results in more change-overs, and here is where most people get hung up on. "Change-overs are bad", they say... and I agree, unless I have capacity left and achieve smaller lot sizes with it.

What I am saying here is, that you are better off using superfluous capacity on the line with change-overs rather than making product that you don't sell or shutting down the line. Because the increase in change-overs comes from a minimizing of lot sizes, which in turn, gives you much more flexibility and helps you move towards a more (actual) demand driven supply chain.

And here is the kicker... I see companies spending an incredible amount of time breaking their heads over optimizing change-overs or setups or production programs. All these Advanced Planning systems market themselves with heuristics and optimization strategies that help you fill the line, so that you can fill the line with waste! You say no? as soon as I produce anything over demand, I produce waste. And that happens when I either fill extra capacity with product that I don't need or when I produce in big batch sizes and have no way to undo that big production run from Monday morning through Tuesday afternoon, when on Thursday, it turns out that I really didn't need that much.

Besides... here is a little anecdote about determining the best possible schedule: If I have one work center and I need to schedule three jobs I have 3! possibilities. Therefore I have 3 times 2 times 1 possibilities which is 6 (first A then B then C or first B then C then A or...). However if I have one work center and 25 jobs to schedule on it, I have 25! possibilities. That is a number when put in pennies, it covers the state of Texas... 6 miles high! How many of you have more than one work center? and what heuristic can produce a good result for that?