
Intelligent Systems Report September 1996 Volume
13 No. 9
The virtual reality (VR) field suffered a potentially crippling blow
earlier this year when its largest conference and exhibition - the
VRWorld show - was abruptly canceled. Fortunately, the VR industry
had an even bigger venue at which to showcase its latest
technological marvels - the huge SIGGRAPH '96 Conference on Computer
Graphics and Interactive Techniques, held last month in New Orleans,
La.
Certainly not all of the 28,500 SIGGRAPH attendees ventured to the
Mississippi Delta solely to see the state of the art in VR
developments, but thanks particularly to the emergence of the Virtual
Reality Markup Language (VRML) 2.0 standard, VR's influence was
ever-present and profound.
All of the energy and enthusiasm that used to characterize the AI
field in the 1980s seems to have shifted over to such pursuits as 3-D
interactive design, visual simulation and virtual prototyping.
Conference chair John Fujii of Hewlett-Packard opened the technical
program by observing, "By taking a chance here, by seizing an
opportunity there, [we] expand the field of possibilities. It is not
merely the power of the processor, but it is the power of the ideas
springing from our connected pursuits that guarantees an exciting
future for us all." And indeed, with more than 320 exhibitor booths
luring attendees away from the technical sessions, there was a
plethora of ideas and pursuits, all charged by the opportunity to
further the development of computer graphics.
In this special SIGGRAPH '96 show issue of ISR, we will look at some
of the significant VR developments being undertaken by the R&D
community, as well as profile the latest generation of hardware and
software introduced at the exhibition.
HMDs vs. SIDs
SIGGRAPH '96 devoted a panel presentation to forecasting the future
of virtual reality, centering on a debate of the merits of the
traditional head-mounted displays (HMDs) versus the more recent
spatially immersive displays (SIDs) such as the University of
Illinois' CAVE, a theater-like domed video projection.
According to VIRTOOLS' Bertrand de La Chapelle, "HMDs are ill adapted
for day-to-day professional applications," due to such problems as
weight, eye and neck fatigue; the isolation factor that makes it
difficult for collaboration; the need for better performance; and
even the psychological stumbling block that HMDs are considered game
gear and are not taken seriously in business environments. He
believes that SIDs offer a better chance for collaborative
applications, though they too face major drawbacks: underdeveloped,
very expensive and requiring much more space and hardware.
"As VR applications evolve from simple walk-throughs to virtual
working environments, SIDs might become the new paradigm for
professional use," De La Chapelle predicted. "Key developments in
graphics hardware and projection devices will create a range of
standard systems, from individual large screen displays to
full-fledged multi-participant domed environments."
Alternate Realities' David Bennett agreed that spatial immersion,
particularly dome projection, "provides a better solution than other
alternatives for groups of people." While HMDs remain a valid and
important development environment for single users, Bennett observed
that "acceptance of domes for VR is greater than for HMDs since it
does not require wearing any restrictive devices and `feels' more
like what is expected of VR."
Mark Bolas of Fakespace cautioned against abandoning the HMD simply
because of its technological challenges. "Are we so lazy that we'll
shy away from the hard problems of cutting-edge immersive displays to
move toward the relative utility and security found in
projection-based and other alternative displays?" According to Bolas,
the R&D community needs to focus on making "both the display
technology and the content work together to form a seamless immersive
experience."
The emergence of VRML
VRML 2.0, also known as Moving Worlds, has emerged as the standard
multimedia description language, enabling interactive 3-D animations
for the World Wide Web. Thanks to VRML 2.0, "Networked,
interoperable, interactive 3-D graphics are now feasible for
everyone's computer," remarked Don Brutzman with the U.S. Naval
Postgraduate School.
"Although networking has been considered `different' than computer
graphics, network considerations are integral to large-scale
interactive 3-D graphics," he observed. "Graphics and networks are
now two interlocking halves of a greater whole: distributed virtual
environments. Our ultimate goal is to use networked interactive 3-D
graphics to take full advantage of all computation, content and
people resources available on the Internet."
A lofty goal indeed, and as Silicon Graphics' Gavin Bell explained,
VRML 2.0 has been viewed as successful because "it hasn't tried to
solve all of the problems of computer graphics, simulation and
networking all at once." As one of the architects of VRML 2.0, Gavin
described what work remains to be done: "Tackling the multi-user
problem will require agreement on exactly which problem should be
solved - multi-user chat is a much easier problem than general
multi-user collaboration in a shared virtual world. Solving the more
general problem will require additions to both VRML and the
infrastructure of the World Wide Web."
Will workstations become obsolete?
With the emergence of more powerful and affordable graphics
accelerators for PCs, some pundits have begun to forecast the
eventual demise of workstations. For instance, Michael Cox of S3
predicted, "PC graphics will drive workstations into Chapter 11. At
the top of the market pyramid are the few users who need astronomical
performance at astronomical prices. At the bottom are the PC masses.
PCs will ultimately take the whole pyramid because of unyielding
forces: business model, technical requirements, and volume."
Not surprisingly, Sun Microsystems' Michael Deering sees the future
quite a bit differently. "Three-D graphics hardware for PCs have
finally found their killer apps: home games. The most successful PC
vendors are busily dumping their half-hearted attempts to achieve
workstation quality in an effort to shave a little cost off to be
more price competitive in the cutthroat 3-D home game market."
While vendors will continue to expound on the state of the industry
as it reflects most favorably on their own niche, the undeniable
truth that emerged from SIGGRAPH '96 is that whether it's on a PC or
on a workstation, whether it's delivered via an HMD or within an SID,
whether it conforms to the VRML 2.0 standard or not, virtual reality
has become a mature technology for developing and enhancing
intelligent applications.
Click
here to return to Table of Contents for the ISR September
issue