Thursday, November 14, 2002



George Koenigsaecker of Simpler Consulting proposes 9 ways Lean is different from our "natural" way to think. Here's his point 4, in the form of "natural vs Lean."

Economic Order Quantity vs. All set ups can be reduced by 97%

This of course is a further variation on the batch vs flow discussion of last week. EOQ recognizes large fixed costs as "normal". An effort towards set-up reduction says "we don't have to assume that large fixed costs are normal." A couple of examples.

I had a tour about 18 months ago of a large metal extrusion facility. They had a huge (and I mean huge) 800 ton extrusion unit to squeeze out piles of metal to customer-requested profiles. The kicker was temperature control. Every time they had a new shape requested, they had to swap out the die and wait for it to heat up. "Conventional wisdom" said this was a minimum of an 8 hour process; during which nothing happened. The fully staffed team sat around and played cards, just waiting for the die to heat up. As a result, whenever the die did heat up, they made the same part for several days on end...they simply could not afford to shut it down after only a few parts.

In a process overhaul, managers challenged this assumption and provided training on quick change-over. Two main things happened.

  1. They paid attention to the schedule and knew if Die A was in the machine now, the next one required would be Die B.
  2. They installed a pre-heating oven. While Die A was running, Die B was warming up to the key temperature. They then put a "hot" new die in the machine when ready.
  3. They tested just how critical the temperature was to the metal. Turns out the exact temperature was a theoretical issue but not a practical one. The metalurgists had insisted on a level of control that just wasn't necessary.
As a result, they dropped the changeover time to less than 12 minutes. Yes, from 480 minutes to 12 minutes. I saw it. How? By challenging assumptions.

I'll have another example tomorrow.

Feel free to forward to a friend.


Email me



No comments: