Would you say a parallel to the example below, that having terminology
and translations automatically generated...would be like automatically
generating an ontology's "geometry"? Auto-syntax? (01)
On 2/22/07, John F. Sowa <sowa@xxxxxxxxxxx> wrote:
> I just wanted to add an observation that is related to
> the questions about ambiguities in natural language and
> the critical need for precision in certain applications.
>
> There was a recent story about problems in the development
> of the Airbus 380. One of the critical issues was that
> the 300 miles of wiring that was being assembled in France
> didn't fit through the holes in the fuselage that was being
> built in Germany.
>
> The Airbus and Boeing designers were using the same CAD
> (Computer Aided Design) system called Catia. But Boeing
> forced all their developers and suppliers to upgrade all
> their software in lock-step every 3 months.
>
> For Airbus, however, the French were using Catia Version 5,
> but the Germans were using Catia Version 4. The result was
> that the holes designed by the Germans didn't fit the wiring
> designed by the French. (See below.)
>
> In English (and every other NL), there is a word for "hole",
> but the exact details of how a hole is specified are seldom
> stated precisely. But for designing an airplane the amount
> of detail can be staggering. Discrepancies between the way
> a hole is specified in the two systems created an ambiguity
> that was not detected during the design stage.
>
> They used to say that the complete specifications (on paper)
> for the Boeing 747 wouldn't fit in a 747, and if it could,
> it would make the plane too heavy to take off. But the amount
> of detail for the new planes is so much more voluminous that
> nobody ever puts it all on paper. For Boeing's forthcoming
> 787 (the Dreamliner), the amount of data for the design takes
> about 16 terabytes of storage.
>
> That's a lot of potential for unnoticed ambiguities to create
> disastrous problems -- which can, as in this case, cost
> billions of dollars in redesign and lost sales. Even worse,
> some problems could go undetected until the plane is built and
> put into service. Then it becomes a matter of life or death.
>
> John
> _____________________________________________________________
>
> Both [versions of Catia] are able to represent objects in 3D, but that's
> where the similarities end. Engineers using Catia V4 must use a manual
> process to create the geometry of a model. To create a hole inside an
> object, for example, the system requires them to subtract a cylinder
> from the space to define where the hole should exist. By contrast, the
> product designer using Catia V5 simply feeds in a set of engineering
> instructions -- in effect, describing the location and dimensions of the
> hole -- and the geometry is automatically created. "V5 is higher-level,
> more intuitive," says Doug Cheney, product manager for CAD
> interoperability quality at ITI TranscenData, a developer of CAD
> translation software. "With the older system, the engineer figures out
> the geometry; with the new one, the system finds the best geometric
> solution."
>
> Airbus engineers ran afoul of this basic difference when creating the
> miles of wiring to be inserted inside the A380 fuselage. The engineers'
> "notes" -- appendices that describe details of models such as curves --
> sometimes are not replicated in the translation between Catia V4 and V5,
> says David Prawel, president of Longview Advisors, a Loveland,
> Colo.-based consulting firm specializing in 3D software issues for
> manufacturing. In other words, key notes required to duplicate a 3D
> model showing electrical wires as they twist and bend through the
> aircraft may fail to reappear in full and accurate detail when a design
> file in one system is converted to a file in the other.
>
> Source:
>
> http://www.baselinemag.com/print_article2/0,1217,a=200193,00.asp
> PLM: Boeing's Dream, Airbus' Nightmare
>
>
> _________________________________________________________________
> Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
> Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
> Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
> Shared Files: http://ontolog.cim3.net/file/
> Community Wiki: http://ontolog.cim3.net/wiki/
> To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>
> (02)
-- (03)
************************************************* (04)
Deborah MacPherson
www.accuracyandaesthetics.com
www.deborahmacpherson.com (05)
The content of this email may contain private
confidential information. Do not forward, copy,
share, or otherwise distribute without explicit
written permission from all correspondents. (06)
************************************************** (07)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (08)
|