home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.engr
- Path: sparky!uunet!utcsri!skule.ecf!drill.me!fil
- From: fil@me.utoronto.ca (Filippo A. Salustri)
- Subject: Re: Can engineering be automated?
- Message-ID: <Bs5yu4.CDK@me.utoronto.ca>
- Sender: news@me.utoronto.ca (News Reader)
- Organisation: U of Toronto, Dept. of Mechanical Engineering
- Organization: UofT Mechanical Engineering
- References: <1992Jul20.174628.25417@cbfsb.cb.att.com> <19570003@hpfcso.FC.HP.COM> <1992Jul29.125652.13819@cbfsb.cb.att.com>
- Date: Wed, 29 Jul 1992 18:46:51 GMT
- Lines: 118
-
- rizzo@cbnewsf.cb.att.com (anthony.r.rizzo) writes:
- >[...]
- >But these tasks, by themselves, are not engineering.
- >The engineering takes place when the calculations are automated, not after.
-
- I *could* argue that the automation of calculations is computer
- science, and not engineering. However, I think it would be more accurate to
- say its `software design'. Whether this is, or should be, a component of
- engineering is open to debate. Personally, I think it should be, but that's
- another topic. I would say that a more precise statement is: ``Design (of
- software) takes place when the calculations are automated''. As far as I
- can tell, `engineering' is bigger (or just more vaguely defined?) than
- `design'.
-
- >The engineering takes place when the parts-list software is created, not after.
-
- Ditto.
-
- >The engineering takes place when the information is first obtained, not
- > when it is cataloged.
-
- The process of cataloging, as I understand it, is an organizational
- one requiring a good understanding of the intended use of the information.
- This strikes me as a `design' problem. For longer (and better) arguments in
- favor of this point of view, I would suggest Suh's book, "The Principles of
- Design", Oxford Press.
-
- >The engineering takes place when the algorithms are invented, not after
- > they're programmed.
-
- Invention of algorithms sounds like computer science. Perhaps you
- meant `methodologies'? (This is a term generally accepted in mechancial
- engineering, with which I am most familiar). Methodologies can be defined
- as processes by which classes of problems can be solved. The reason I pick
- on this is that we (mechanical engineers) come up with methodologies first,
- then create `algorithms' when we want to implement software. It's just a
- matter of terminology, and I would like to make sure I understand you
- correctly.
-
- >The engineering is in knowing which of the automated calculations
- > are appropriate.
-
- This is part of the general problem solving process. Insofar as
- that goes, I agree; alot (maybe all?) of what engineering is relates to
- problem solving. This would be a good area for someone to start doing
- research. Given a number of automated calculations, and a problem which is
- to be solved using one of the calculations, what kind of `knowledge', or
- maybe just information, is needed for a confident decision to be made? The
- problem can be narrowed by restricting the kinds of automated calculations
- to those types of calculations typical in engineering or some component of
- engineering. If we can build a formal model to capture this, then we will
- be able to confidently say that this segment of engineering could be
- automated too (i.e. a formal model could at least theoretically be
- implemented in a software system). If no formal model can be found, then
- barring massive heuristic systems, we will be able to confidently say that
- this segment of engineering cannot be automated.
-
- The important point here is that it is possible to determine with
- some confidence whether some segment of engineering can or cannot be
- automated, and that opinion is not sufficient; formal means do exist to make
- this determination.
-
- >The engineering is in knowing what part of the geometry is sufficient
- > to perform the automated calculations.
-
- Ditto.
-
- >The engineering is in modifying the already programmed algorithm or
- > in using it in a new and creative way to solve a NEW problem.
-
- This is `redesign', which is strongly related to `design'.
-
- >The engineering takes place every time we use a calculator instead of a cray,
- > because the calculator is sufficient.
- >The engineering takes place every time we use a cray instead of a calculator,
- > because the solution REQUIRES a cray.
-
- This is not engineering; this is common sense. Although it is
- *highly* beneficial for engineers to have common sense, it is a not
- particular to engineers or engineering and so is related to it only
- peripherally.
-
- >The engineering takes place every time we look at computer-generated
- > numbers and INTERPRET them as either a solution or as garbage.
-
- Interpretation of data is by no means only possible by humans.
- Programs interpret data all the time. You are reading these characters only
- because various programs are interpreting a bunch of data as characters.
- Granted that programs are not capable of the same degree of interpretation
- as human brains are, but it is only a matter of degree.
-
- >There you have my reasons for saying "NO!" I'm sure that there are many
- >who don't agree. To them I say this. May your family fly in a plane
- >that was designed by competent engineers and not by technicians with
- >automated tools.
-
- I assume this is meant to wish me well, though I could argue for the
- opposite `interpretation' also. :-)
-
- I've presented some counter-arguments here not necessarily to pick
- on you, Anthony, but rather to indicate that the issue is by no means as
- simple as some people may think. I believe it is possible, though, to reach
- a good and reasonable conclusion through the kind of discussions that have
- taken place here so far. So we're off to a good start.
-
- > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
- > * Anthony R. Rizzo | *
- > * The FEA Group | Latest republican campaign slogan: *
- > * AT&T Bell Laboratories | *
- > * att.com!homxc!rizzo | "George Bush, only four more years!" *
- > * (201) 386-2565 | *
- > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
-
- Cheers.
- Fil Salustri (fil@me.utoronto.ca)
- --
- Fil Salustri fil@me.utoronto.ca UUCP: ...!utai!me!fil
- A is A
-