A MAN AND A DOG

A few months ago The Atlantic had an interesting article about a manufacturing plant in South Carolina (http://www.theatlantic.com/magazine/archive/2012/01/making-it-in-america/8844/). The EconTalk podcast had an equally interesting interview with the author that I was listening to recently. (http://files.libertyfund.org/econtalk/y2012/Davidsonmanufacturing.mp3)

They repeated something that I have heard before but had not thought much about:    ”In a highly automated manufacturing plant the labor force consists of a man and a dog. The man is there to feed the dog. The dog is there to bite the man if he tries to adjust a machine.”

I’ve spent a lot of time in manufacturing plants over the past 35+ years. Much of that has been spent with automated packaging, manufacturing and assembly machinery. A tendency I often see (and am guilty of myself occasionally) is what I call “tinkering”. This tinkering is what the dog is there to prevent in the above jape.

The machine will be chugging along making “good” ie; non-rejected, products at normal speeds. They think that maybe if they just tweak the temperature, loosen or tighten that guide rail, adjust the clutch pressure or whatever it will work even better. It seldom does. Quite often this tinkering makes the machine run worse. There are several reasons that this happens.

One is that most people honestly want to make the machine run better. Another, I believe but nobody wants to admit, is that they get bored just watching the machine run and want to do something.

A third factor is that they often do not understand the machine. They likely understand the mechanics of it but may not understand the theory behind the mechanics. They understand the “how” without understanding the “why”. Hence they make adjustments without understanding all the implications and possible effects.

Finally they may not understand the concept of normal variability. All processes will vary. We should always be working toward eliminating this variability, though we will never get there. This is only accomplished by improving the process, not by constantly adjusting.

Occasionally tinkering will result in improvement. In most cases, it is at best waste of time, at worst it will result in a machine running more poorly.

Tinkering must be prevented. Here are a couple ways to help prevent it:

Training – All teammates on the line, whatever their job is, must understand not just what they are supposed to do but why it is to be done that way. They must be trained in the theory behind the machines they work with. They must be trained to stop and think before they make any change to assure that it is justified.

Process capability – Everyone on the packaging floor (as well as others off the floor) must understand the capability of the process. A liquid filler may have a capability to fill to +/- 1 oz. This may be due to filler design, product, filler wear or other causes. If the teammates don’t understand this, they may be constantly trying to keep it within +/- 1/2 oz. This is a losing game that will result in lost efficiency and frustrated teammates. The only way to achieve +/- 0.5 oz is to improve the machine capability.

Supervision – You probably don’t want a dog on the packaging floor (especially in a food plant). You do have a supervisor on the floor. That supervisor must be taught to keep an eye out for unnecessary tinkering. They must stop it when they see it.

Stability and consistency are the keys to efficiency. Figure out how machines are supposed to be set and how they are supposed to run and stick with it.