From: mbernat@gel.ulaval.ca (Marc Bernatchez)
Subject: Re: INDUSTRY:  Successful VR Apps
Date: Tue, 03 Dec 96 15:37:12 GMT
Organization: Laval University


>My question:  Does anybody really use VR for real work?
>In other words, I'm interested in hearing what people consider
>examples of "Successful VR applications".  Specifically, I am looking
>for examples of applications that have been implemented in a virtual
>environment that have demonstrated some measure of success
>(commercial, solved a problem that couldn't be solved otherwise, etc).
>
<...deletion...>
>   -  Architectural walkthrough (such as the work here at UNC)
>   -  Phobia treatment via exposure therapy (like the work at Georgia Tech)
>   -  Scientific visualization (NASA's virtual windtunnel, UNC's
>      nanomanipulator project)
>   -  Marketing (virtual showrooms)

Yes, there are some very successful applications in these branches of
development. Especially the ones at NASA. I don't know if the wind
tunnel is commercially used but it's well done.

Yet a best example of success, their desktop VR system they used to
fix the Hubble telescope. It really saved the day. They had forgot a
flaw in the design and found it soon enough with the VR
simulation. Without it, they could have shipped a flawed fixing
apparatus to the sky and find the problem only when swapping the
mirror elements on the telescope...hard to begin fixing such precision
devices with the astronaut suit!

There is also the design of the 777 Boeing plane that was built a bit
every where on the planet by separate teams... their only link was a
CAD/pseudo-VR(don't know how much VR it was) system that let them
virtually put it all in one piece all along the development time.

Car manufacturer uses VR to build virtual prototypes of the car body
to test the ergonomics and visibility inside the car by example. A
prototype may cost several 100 000$. It's always fun to be able to
build over 10 prototypes and iterate until you aren't satisfied by the
design. In VR, it doesn't cost that much each time. They only build
the final prototype car body (the final mould).

I'll add my own VR system to the list even if it's not as major as
these above. We have realized a VR simulation (desktop based also)
that let engineer walk in a lab room where they can experiment on a VR
representation of a real lab setup. They can do the same
manipulations, setting parameters on the instruments (scopes, waveform
generator and so on).

I'm sure Linda Jacobson can add to the list if she read this ;-)

>I'm most interested in hearing about immersive virtual environment
>applications using head-mounted displays or Booms (my personal bias),
>but I'd also be interested in hearing about "Cave" and "fishtank VR"
>type applications as well.

These technologies are still much in labs development. Most major
successful VR system aren't using these fancy visualization method
yet. The CAVE may be used a bit though. Most current successful VR
systems are desktop based. Some may use HMDs but in a limited way
(depending if the visual resolution is important or not in the
application... and it usualy is).

>In forming your response please indicate how much you think the
>"success" of the application is due to being in a virtual environment
>(immersion, direct manipulation, etc.) and how much is just due to
>real-time interactive graphics (i.e. it would have worked just as well
>sitting in front of a conventional computer monitor, forget all the
>fancy tracking, displays, and input devices).  Also, what is keeping
>the "near misses" from being "direct hits"?  Cost?  Quality of the
>technology?  Wrong applications?

For the Hubble repair...definite yes, the VR part of the system was
necessary and saved a lot of money.

For the 777, the VR part was also pretty necessary. Try to build a
jumbo jet class airplane with an under the millimeter precision given
you have spread teams all over the planet working on their side.

The VR side of it is helping a lot car manufacturer in their
design. They can fly through the car before it's been concretely made
and "see" what it will be like to be in this given car.

Our own system is still in development but we have already got good
comments about it. It has been found to be easier to do complex lab
manipulation in that VR environment than in the real life lab
setup. The VR GUI I have designed also greatly help people interact in
such worlds. An other benefit of such a VR lab is that you can
actually provide more content than in the real life situation. I can
show dynamically updated theoretical graphs along with the rest of the
lab setup. In VR it's no problem, I can make that graph float in air
above an instrument...I couldn't with the real thing.

So you see, there is a lot more done with VR than we think. It's still
very young as a technology or science even, but it's coming with giant
steps. The 3D accelerator cards and other such technologies are
entering the mass market arena and it won't be long before we can
build VR system that are efficient and cheap.

Later

 Marc Bernatchez             |  E-mail: mbernat@gel.ulaval.ca
 C.O.P.L. (local plt-00307)  |============================================    
 Dep. Genie electrique       |  http://www.gel.ulaval.ca/~mbernat
 Universite LAVAL            |  http://www.imaginative.com/VResources
 Quebec, Canada              |============================================
 G1K 7P4                     | Virtual Reality is the future of computers
==========================================================================