Shift from craftsmanshift towards engineering in manufacturing? 5

[MartinLe]

I have a hunch: Until a few decades ago, engineers would do their design, then in many cases the foremen or other skilled crafts persons on the shop floor would figure out what jigs to build etc. to manufacture something efficiently. Anecdotally this is confirmed by stories from one of my grandparents who was a master machinist/turner.

Then, with the arrival of CN machining going on to current robotics, the relative importance of master craftsmen (well, mostly men) etc. in the manufacturing waned, more of the actual manufacturing process was designed off the shop floor.

My question is if this hunch is basically or at least somehwat correct?

Mind, I recently read one article about modern chip manufacturing: The large foundrys (the example was in Taiwan) employ armies of (mostly young women) to do all sorts of manual tasks in the clean area, working 12 hour shifts without lunch breaks, because the production runs change so fast that they can't reprogram or redesign all the robots. Another recent article I read about a solar cell manufacture made the point that the US lead in research but that the technological advantage Japan gained stemmed from actually manufacturing PV cells (for wristwatches etc. at first)- the claim is essentiall shop floor knowledge helped secure the advantage. So only a relative shift, not an absolute one.

 

[JohnRBaker]

Note that comment about the Japanese, this brings me back to when we first started to do business in Japan.

Back in the early 80's we signed a deal with Seiko for them to be our partner selling and supporting our CAD/CAE/CAM software in Japan. In those days the only practical way for an American company (I was working for McDonnell Douglas) to do business in Japan was to partner with a Japanese company. Note that Seiko was already using our software in a few parts of their company (mostly tooling for manufacturing watches), and of course, once they were our sales partners, they were able to procure more seats of our software at much better prices. Now getting to the point I'm trying to make, after a couple of years of Seiko selling our software we were starting to have problems with their customers NOT paying maintenance fees.

For those not familiar with how software was sold back in the 20th Century, we licensed our software on a per-seat basis. Now the software license was in perpetuity, but if you didn't pay a monthly 'maintenance fee', you didn't get things like hotline support, bug fixes nor updates/enhancements. Now back then these 'fees' ran around 1.0% of the retail price of the software per month (years later this went up to 1.5%/month). Now this might sound high, but for this you also got upgrades, that is all the enhancements that R&D had been working on, in addition to bug fixes. Our software was very complex and while we started out sending 'patches' to fix bugs, and we still did for years, most fixes came when we updated the software, usually every 12 to 18 months, and with those updates you'd get all the fixes as well as improvements and enhancements which we had incorporated in the software since the last release. We only sold ONE version of the software, granted, it was ported to many different hardware platforms, but functionality wise, these 'versions' were identical.

Now the problem with Seiko was that if their customer felt that there were bugs in the code, even if they were not show-stoppers (they had acceptable workarounds), the Seiko people just felt that it was a poor business practice to continue to collect the monthly maintenance fees. In fact, I had one of the Seiko executives tell me that if we could deliver a version of the software with absolutely NO bugs, that their customers would probably be willing to pay even more maintenance than the 1.0%/month. We tried to explain to them that there's no such thing as bug free software. The old adage was that if there were X number of lines of code in a routine then there was X+1 number of potential bugs (now things were NEVER that bad, just that bug-free code was very rare). Of course, their argument was that their customers expected perfection and they were willing to pay for it, when they got it, but not if they didn't. We eventually overcame that cultural hurdle but it took years.

John R. Baker, P.E. (ret)
Irvine, CA
Siemens PLM:
UG/NX Museum:

The secret of life is not finding someone to live with
It's finding someone you can't live without

 

[GregLocock]

To get back to the original point, at least with cars, tooling was designed by the draftsmen. If you have a 2 ton stamping die for a door Jim Bob and Bill in the press shop didn't just start chipping away at the tool steel with a cold chisel.

It is true that if I wanted a one off casting I'd probably walk into a foundry and they'd make the the pattern and the runners and so on, on the fly, but if it was for mass production you'd have it all drawn up, at least in the last 70 years if not longer.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?