(Continued from “Lean Journeys – Part 1“)
There were two departments in the plant, and the press area served both with most of their raw material. After about two weeks on the job, I found that my area was shutting down one assembly cell per day on average. As you could imagine, my stress level was getting very high in a short time. I started making daily schedules of what I wanted to run and when I wanted to run it. I tried scheduling by piece count and by run time, but nothing I did worked. I knew that I had considerable downtime, and my changeover times were considerable, but I was in a constant state of firefighting for weeks.
After consulting many references, I started to take a six-sigma approach to the dilemma. What were the inputs? Obviously changeover time was an input. Running downtime was another….but what was the “Y” here? I initially went with assembly stock-outs per day, since if I could keep the plant running, I could buy time to make the press area more efficient.
First, I had to figure out what my average customer demand was (takt time), and weather the demand was leveled. Leveling of demand is very important, since this makes the production signals of a pull system stable and predictable. As it turned out, my highest demand product was shipped as an assembly in one week quantities to Europe by boat with a two-week lead time….not good, but I considered it system noise since there was no way to immediately control that factor. Luckily, the weekly demand for the product was stable.
As it turned out, taking a hard look at the operations, I found both good news and bad news.
The good news:
I had more capacity than I thought – on average I could run more pieces than my demand required.
The bad news:
1) My changeover times were astronomical.
2) My unplanned downtime was also too high.
What was happening was that the min-max levels for the pull systems were so low, that the average press was almost spending more time changing over than actually running parts!This was causing sheer chaos in the operation.
The improvement plan consisted of:
1) Re-setting min-max levels to reflect operations characteristics (aka â??Take the initial hitâ?).
2) Improve changeover times/reduce unplanned downtime using conventional problem solving.
3) Lower min-max levels to the improved operations characteristics.
4) Repeat the problem-solving process.
My next blog will discuss the before and after results of the activity, and will also discuss some of the conventional stumbling blocks for going lean.
Please take a look at the following as they are really excellent references:
Kaizen for Quick Changeover: Going Beyond Smed by Kenichi Sekine, Keisuke Arai, and Bruce Talbot
Factory Physics by Wallace Hopp and Mark Spearman