RE: Article in CIO magazine

From: Charles Overy (charles@lgmmodel.com)
Date: Sat Oct 05 2002 - 03:46:55 EEST


David,

Thanks for your reply.

I stand corrected on the AutoCad on Mac. I am a tremendous supporter of
Apple and was disheartened when we had to switch to Windows based machines
about six years ago as we transitioned into 3D, machining and now to rapid
prototyping. I wish it was currently more feasible to have Macs in our
business.

I find it amusing, although I am sure that AutoDesk would not, that you
intended "AutoCAD" to mean "any old 2D program". I did not make that
connection.

I too would love to see even a demo of a program that will nearly
autonomously generate a 3D model from 2d sketches, particularly one that can
subsequently generate buildable .stl files. It would be very very useful
in our work. I would agree that there are a lot of tools that have some of
this utility if you do already have sufficiently accurate 2D CAD drawn from
enough perspectives to define the object. There are in fact tools to
generate 3D from 2d architectural drawings in Autocad. Like a lot of the
photogrametic applications that will generate 3d models from photographs,
the programs that I have tried only work well in very ideal cases. There
does not yet seem to be suitably powerful programming to replace human
intuition, particularly in the interpretation of spatial data.

However, the first assumption for the use of these 2d to 3d products is
that your work process has accurate 2D CAD drafting and you now want 3D from
that data. To those of us already helping people to use rapid prototyping
for early concept design, that assumption negates a great deal of the
potential benefits of RP. Your article seems to suggest, although I agree
it is not a major tenet, that 3D objects might still be better designed in
2d. Personally, I think that the cost, features and utility of the entire
3D design process are now at a point that that is no longer true. Companies
like your shoe client perhaps were burned as early adopters of 3d and have
decided, either by design or by default to be less aggressive in the future.
I think they will find this to their detriment particularly in a global
marketplace where domestic manufacturers must appropriately leverage
technology against the high cost of labor. Processes like RP have
traditionally been sold on the concept of cost avoidance or time to market
in high production, design once, build many applications. They have proven
their worth here. I would assert that the technologies are now suitably
mature so that one can at least make a sound argument that 3D objects are
better designed in 3D. Yes, from sketch to delivery, the adoption of 3d
will, in the aggregate, lead to better more innovative products.

I would be happy to talk to your about the details of the various RP
processes but there are better people on the RP-ML list than me. (pass the
buck...) I would imagine that you received a number of links of various
sources describing the processes, if not I will dig some up.

I would agree that is only modestly comforting that in 100 or so years of
technological progress we are now comparing the death of John Henry to an
all nighter in a Chinese sweat shop. Given a choice between the two I would
take the latter although not by much. In today's uncertain times, we
probably need to praise the small advancements in the human condition when
we can. Sometimes I fear they might not last.

Best wishes.

Charles

PS. I don't really think you are a Luddite, but some of my clients are.
Chalk that one up to Friday afternoon. Someone in the office brought me a
beer and I feel much better now.

-----Original Message-----
From: David Dobrin [mailto:david@b2banalysts.com]
Sent: Friday, October 04, 2002 3:34 PM
To: cwho@lgmmodel.com; rp-ml@rapid.lpt.fi
Cc: dduffy@cio.com; rpastore@cio.com
Subject: RE: Article in CIO magazine

Thanks for the help with this. I am myself a stickler for accuracy and
always want to make sure that I get everything correct, so I appreciate this
e-mail, whatever its combustion temperature, very much.

The events described in the article were accurately described. The
terminology was the terminology that was used to me by people who are in a
position to know, so I didn't question it.

STL is often used as a generic abbreviation for stereolithography (check the
Internet), but I agree that it is only in the context of a file format. I
should probably have used SLA or rapid prototyping machine. I think what
happened is that I got caught up in the question of whether we can generate
an STL file, and if so, how. For me, an STL machine was therefore one that
can print an STL file.

In my own defense, I might argue that my usage was not so much incorrect as
unidiomatic. Neither, however, meets the ideal that you and I both have for
any article. Thanks for the catch.

As far as the more substantive comments, let me take up the science fiction
question first. I have had a good relationship with the people at
IBM/Dassault for several years. When we first bruited the question of taking
a shoe design in 2-D and printing it out on an SLA machine, they were so
willing to persuade me of their capabilities in this area that they set up a
2-hour conference call with a Dassault technical expert. For simplicity's
sake, I used AutoCAD as an example, presented him with the problem, and let
him noodle about it. He said that they do in fact have programs to convert
multiple 2D views to a single 3-D. He said that theoretically, the drawings
should convert perfectly, but that slight differences in specifications
among the views might require some manual labor in order to get a single,
consistent 3-D view. The tweaking is really just reconciling these slight
differences.

They were willing to do a demo (2D-->3D) for us on our 2-D AutoCAD sketches,
but (as you saw), we couldn't even get 2D drawings of a shoe. If what they
told me is science fiction, well, I was snookered. But I was not credulous.
I spent a lot of time trying to understand exactly how this was done, and
what he told me was detailed and consistent. If it were to turn out that you
were wrong and that Catia can in fact do what Dassault/IBM says it can do,
don't you think you should reconsider the question of whether the article
belongs in People rather than CIO?

As far as the AutoCAD for the Mac is concerned, that may have been a little
fey on my part, or I may have just missed the audience's capabilities. I
wanted to convey our thought processes, and since both the CIO and I think
in terms of 2D or 3D and think of AutoCAD as the generic 2D drawing program
(and yes, I know it has 3D capabilities), I just used that as a stand-in for
any old 2D program. I figured my audience would not know or care whether it
was AutoCAD or some other 2D program that the designers were(n't) using over
there.

AutoCAD on the Mac has had a checkered history. (See, for instance,
www.architecturalcadd.com/autocadmac.html, which just pops up on Google,
when I put in AutoCAD, Mac into the search engine.) But it is surely
incorrect to say that no AutoCAD product has ever run on a Mac.

You also say that my description of the rapid prototyping process is
incorrect. That may be. The (color) shoe that was plopped on the CEO's desk
was made by a Z corporation machine, I believe. The explanation of how these
machines worked was given to me by people at 3D Systems. I may indeed have
missed some nuances in the way various different machines work. All I can
say is that I was abetted by Dassault, who told me that they could output
the finished file to any of the rapid prototyping machines and that the
capabilities were roughly similar.

Naturally, I would love to learn what those nuances are; if you feel you
have the time to educate me, rather than scorch me, please feel free to give
me a call any time.

Thanks, by the way, for appreciating some of my nuances, even as I missed
some of yours. I was indeed hoping to suggest that there may be something
problematic about the labor practices (and customs practices) in China. But
remember, in this case, nobody died, and whatever the cruelty and injustice
to factory workers in China, it would have to be pretty terrible even to
come close to matching what happened to the freed slaves in our own country
100 years ago. In my story, nobody dies, and that's part of the point, too.

It's interesting that you accuse me of being a Luddite. Remember, I'm the
guy who suggested using this technology in the first place. If I had
succeeded, and if the poor Chinese whose jobs I was trying to take away had
been able to get to me, I would have been one of the first targets.

Thanks for cc'ing me on the note.

David N. Dobrin, Ph.D.
President,
B2B Analysts, Inc.
617 497 6399

-----Original Message-----
From: Charles Overy [mailto:cwho@lgmmodel.com]
Sent: Friday, October 04, 2002 4:09 PM
To: rp-ml@rapid.lpt.fi
Cc: david@b2banalysts.com; dduffy@cio.com
Subject: RE: Article in CIO magazine

Hello all,

whoosh -- hear that flame comming at you....

Nice article except for the factual errors! I caught at least two.

"who has so far outperformed a stereolithography (STL) machine. Have you
ever seen one of these? It can take a virtual 3-D image from a software
package such as Catia and put out a colored 3-D model made of plastic. The
STL machine divides the 3-D image into thin layers and makes the model by
laser printing layer after layer of plastic"

 - Not a very accurate description of any RP process that I have seen. It
seems to be an amalgam of the Z Corp process (color), Stratasys (layers of
plastic) , and 3D Systems SLA. Also, since when was the class of machines
abbreviated with the file extension? This laxity maybe OK for casual
journalism but it makes me treat the rest of the article as just that. Not
very well researched or thought out.

"AutoCad and its myriad competitors were surely being used on all those Macs
over in design. "
Again, to my knowledge, no AutoCad product has ever run on a Mac and not
many direct competitors in 3D product design are there either.

Next, can you really, "feed 2-D images into Catia and get an exact rendering
without too much manual tweaking" and if so what does a "rendering" have to
do with a watertight 3D polygon mesh export. If Catia can really create a
good 3D model from 2D hand sketches without "too much" human intervention,
then this is an achievement that I have not heard about. Personally I doubt
it, and it is this sort of statement that is particularly irritating. It
over promises the integration of technology in the design process as a
whole. Further it implies that the RP process get blamed when things break
down. If the authors statement is supposed to be that good people are
needed in the design process, there is no debate bu there are more accurate
ways to convey this obvious truth.

Finally, I think that the RP sales people need to find out who this shoe
company is and leave their card. As the article points out, John Henry DIED
in the process of trying to beat the steam drill. I thought shoe companies
already had enough bad press for inspiring that sort of work environment in
their subcontractors.

This article would have been OK if it had been in "People" or perhaps even
"Gardening News" but in a zine presumably directed at Chief Information
Officers of big companies is should try to get the technology parts close to
correct. I have included the editor and writer in this email, perhaps they
would care to respond to the list. It certainly was an amusing read but I
have more than enough anti-technology drivel from neo-luddites in the
architectural community to last me to the end of this year. Next year,
we'll see...
Best wishes to all

Charles
Dir of Eng.
LGM
www.lgmmodel.com,

-----Original Message-----
From: owner-rp-ml@rapid.lpt.fi [mailto:owner-rp-ml@rapid.lpt.fi]On
Behalf Of Todd Grimm
Sent: Friday, October 04, 2002 12:21 PM
To: rp-ml@rapid.lpt.fi
Subject: Article in CIO magazine

Sometimes it takes an outsider to clearly see what the RP industry faces
as competitive pressures. In a slightly humorous article, CIO magazine
illustrates the challenges that RP and collaboration tools face. Hope
you enjoy it.

http://www.cio.com/archive/100102/worldview.html

Todd Grimm
3028 Beth Ct.
Edgewood, KY 41017
Phone: (859) 331-5340
Fax: (859) 331-5342
Cell: (859) 240-0574
Email: tgrimm1@insightbb.com

For more information about the rp-ml, see http://rapid.lpt.fi/rp-ml/

For more information about the rp-ml, see http://rapid.lpt.fi/rp-ml/



This archive was generated by hypermail 2.1.4 : Tue Jan 21 2003 - 20:14:23 EET