by Tilo Schwarz
There was one group that could almost walk through the Starter Kata as scripted. It was Sam and Doug working on reducing changeover time for machine group AX-3045. They graphed the overall performance by charting the changeover time for each changeover of the last four weeks.
Then they looked at customer demand and the needed performance. After discussing that for a while, they decided not to change the number of changeovers per shift. Initially, Denise tried to push them to increase the number of changeovers, which would enable them to reduce batch size. Realizing their resistance, however, she decided to let them figure it out by themselves once they were under way.
Drawing the block diagram was fairly easy too. Simplified, the block diagram for each changeover had the same steps:
Prepare for changeover
Stop the machine
Remove fixtures and tools
Clean the machine
Mount new fixtures and add tools
Load machine controller program for new part
Test first part at 30% speed
Check quality of first part
If quality is OK, start the machining process at 100% speed, otherwise adjust parameters until quality is within spec.
Transport dismounted fixtures and tools to the cleaning area
Sam and Doug got a little hung up when they tried to measure several process cycles.
“What’s one cycle here?”, Sam asked Denise at one point. “Or at least what are comparable cycles?'' he added.
“You know, two changeovers are never the same, I mean they are for the same part but between parts they are very different. For one part we might need to exchange only three tools, for another maybe 10 or 15. So that makes a huge difference in time. How can we compare cycles in that case?”
They discussed that for a while. Then Denise directed them back to the run chart with the changeover times for the last eight weeks. Studying that again, Sam and Doug realized that there were basically three groups of changeovers. Super long ones, taking 45 minutes or more. Then a few very short ones, around 10 minutes. The majority, however, was between 20 and 30 minutes.
“Hey, Denise, I think we got something”, Doug called on Denise. “Look, there are three groups of changeovers”. He started to explain what they had found.
At the end, Sam presented their conclusion: “Now if we focus on the group with the majority of changeovers, the ones between 20 and around 30 minutes, we should get comparable cycles.”
“Cool”, Denise replied, “why don’t you go and see if you can observe two or three of these changeover times this afternoon and measure cycle time for them?”
Sam and Doug nodded.
Denise added, “once you get back, I guess you’re ready to establish your one-month challenge.” Sam and Doug grabbed their clipboards and made their way to the shopfloor.