UBC Publications

UBC Publications

UBC Publications

Focus 1999

Item Metadata


JSON: focus-1.0115137.json
JSON-LD: focus-1.0115137-ld.json
RDF/XML (Pretty): focus-1.0115137-rdf.xml
RDF/JSON: focus-1.0115137-rdf.json
Turtle: focus-1.0115137-turtle.txt
N-Triples: focus-1.0115137-rdf-ntriples.txt
Original Record: focus-1.0115137-source.json
Full Text

Full Text

DOLPHIN highlights collaboration between CIGSR & industry
CICSR researchers improving
ISE semi-submersible
It's yellow, but a full-scale submarine
it is not. At just 7 feet in length and 10
inches in diameter, the scale-model looks
more like a torpedo. To a group of CICSR
researchers, the scale-model is a dolphin—
the centrepiece of a two-year collaborative
project with International Submarine
Engineering (ISE) of Port Coquitlam, BC.
dolphin stands for Deep Ocean Logging
Platform for Hydrographic Instrumentation
and Navigation. For the past 15 years ISE
has been building and developing this
unmanned, remotely controlled semi-
submersible vessel for ocean surveying.
Evolving demands
ISE has sold a dozen dolphins to
various commercial, scientific and naval
clients, and their evolving requirements
have driven the need for still more vehicle
research and development. To that end, the
company has contracted a group led by
Mechanical Engineering professors Dale
Cherchas and Sander Calisal. They are using
the scale-model dolphin to optimize and
develop control algorithms for the full-size
dolphin's control surfaces.
The control surfaces are the rudder, dive
and stern planes which permit the semi-
submersible to smoothly dive, surface and
manoeuvre. On dolphin, they are controlled automatically. ISE turned to Calisal and
Cherchas in this phase of development
because of their expertise in automatic control
systems, robotics and hydrodynamics.
Calisal has a background in ship hydrodynamics. He is developing equations for
the forces acting on the dolphin, and a
time-representation of these forces suitable
for control algorithms. Cherchas, a CICSR
member, has expertise in system dynamics
and automatic controls, and is developing
dolphin's adaptive, automatic control
Improving performance
"The ultimate goal," says Cherchas, "is
to develop and supply control algorithms
continued on page 2
Spring 1999 Vol. 10, No. 1
Developing intuitive interfaces 3
Exploring complexity 4
Improving real-time systems 5
How the giraffe got its spots 6
Passing Notes 7
CICSR authors 8 A1996 report on the BC information
technology (IT) industry reported at
least 880 unfilled positions. Continued
growth in this important sector of the
economy means even more demand for
skilled professionals.
It is this reality which has led us to
plan a new degree program, the Master
of Software Systems (MSS).
This new degree—a cooperative
effort by CICSR and the departments
of Computer Science and Electrical
and Computer Engineering—is
designed to give students the expertise
needed for the IT workplace.
The program is intended for
students with Bachelor degrees in the
mathematical and physical sciences,
engineering and related areas (other
than computer science and computer
engineering). It will build on the
analytical skills they have acquired in
those disciplines.
The program will benefit both
students and the IT industry; a MSS
degree will give students the software
system skills needed in today's marketplace, and help the provincial IT
industry to expand and innovate with
homegrown talent.
We're very excited about this new
degree, and hope to see the first MSS
students graduate in December of
2000. For more information on the
program turn to page 7, or go to the
CICSR website at www.cicsr.ubc.ca
Rabab Ward, CICSR Director
continued from page 1
that will improve the performance of the
full-scale DOLPHIN."
The CICSR research group—which
consists of Cherchas, Calisal, and Masters
students Adrian Field and Peter
Ostafichuk—is using both computer
simulation and experimental testing to
accomplish this goal.
The full-scale dolphin is propelled by
an air-breathing diesel engine and has a top
speed of 18 knots. As it travels just below the
surface on a route programmed by a remote
command vessel, dolphin transmits survey
data back to the command
vessel/station via a communications telemetry link.
The telemetry antenna is
mounted on the surface-
piercing mast, which also
serves as an air intake for
the engine.
Because dolphin
operates at or near the free
surface, waves and wind
have an affect on its ability to maintain a
steady course as it navigates through its gps-
programmed waypoints. This, in turn, has
bearing on the design of the control surfaces
and the development of the control algorithms.
"We have had to develop a new formulation of fluid dynamics for computer
simulation—a procedure to calculate forces
at or near the surface," says Sander Calisal.
Simulations help measure performance
Computer simulations help the team to
understand the physics of dolphin's motion
through the water. The simulations give
Adrian Field data on the performance of the
vehicle, including the forces and flows
generated by the planes as they respond to
control inputs.
This data helps Peter Ostafichuk as he
works to find the optimal shape, size and
configuration for the control planes. The
different planes and configurations are then
tested experimentally in both a wind tunnel
and a tow tank.
"In a given length of time, we can do ten
times as many tests in a wind tunnel as in a
tow tank," says Peter, "but the tow tank
tests are critical because they include the
influence of the water surface and waves."
Tow tank testing at BC Research
The tow tank tests will be performed at the
Ocean Engineering Centre of BC Research
Inc., the only facility of its kind in western
Canada. The model dolphin will be
equipped with on-board
sensors so that forces and
moments can be measured
during the tow tests.
"The tests will give us
very realistic data," says
Adrian. "As a result, the
control algorithms will be
more realistic."
With about 50% of the
work completed in the
two-year project, the group is now gearing
up for experimental testing, and to that end
has recently ordered a data acquisition
system. Funding for the project comes from
ISE and the BC Science Council.
Hotbed for submarine engineering
Spurred by the growth in undersea
exploration, similar research is also being
conducted at Simon Fraser University and
the University of Victoria. BC is becoming
a centre of ocean engineering expertise, and
the dolphin project has added to that
"The ISE/CICSR partnership is creating
state-of-the-art underwater industry R&D,"
says ISE project director Dr. Mae Seto.
"This type of collaboration complements
the in-house capabilities, objectives and
needs of ISE quite well."
Contact Dale Cherchas at (604) 822-4902,
cherchas@mech.ubc.ca, or Sander Calisal
at (604) 822-2832,calisal@mech.ubc.ca
FOCUS Developing intuitive, adaptive interfaces
Sid Fels works on the next generation of interfaces to ease human-computer interaction
Human computer interaction is commonly
conducted through two interfaces: the
keyboard and the mouse. These devices,
successful as they are, require that we adapt
to their prescribed and limited methods of
use. But what if we had interfaces and input
devices that adapted to our preferences? Sid
Fels believes adaptive computer interfaces
have the potential to make computing more
human and less rigid.
Fels is a CICSR member and a recent
addition to ubc's department of Electrical
and Computer Engineering. He is interested
in creating adaptive interfaces for embodied
systems. Embodied systems are a fusion of
body and computer—a virtual extension of
the body—and require an understanding
not only of software and interface design,
but of human psychology as well.
Adaptive interface technology
"I think that adaptive interface technology can be used to make novel interfaces in
devices such as virtual musical instruments,
and in heavy machinery such as excavators
or cranes," says Fels. "These devices would
adapt to the human controller, making
them easier to learn and use."
Fels began integrating his interests in
computer engineering and cognitive
psychology as a graduate student at the
University of Toronto. His PhD project was
a gesture-to-speech system called the Glove
Talk II. This embodied system translates
hand gestures into speech sounds. Wearing a
data-glove and controlling volume with a
foot pedal, a user can "speak" in a computer-generated voice by combining gestures
which represent parts of speech.
Although the Glove Talk II requires
some training for successful use, its adaptive
nature is noteworthy: the system learns from
the user what gestures are best suited to the
sounds the user wants to make. This
adaptive aspect is key to Fels' thinking
about human-computer interaction.
"I believe that if we can make interfaces
that feel good—that appeal to our emotional sensibility—ease of use and desire to
use them will follow," says Fels. "As we gain
intimate relationships with our machines,
and model them on how we relate to people,
the more these machines will feel like an
extension of who we are."
Intuitive and personal interfaces
Fels sees this intimacy as part of the
appeal of embodied systems, such as the
Glove Talk II, which promise a technology
with more satisfying, intuitive and personal
interfaces. Consider the way people use a
common device like the pen.
"When I hold a pen it becomes an
extension of my hand and is no longer an
external thing—I've embodied it," he says.
"And there is an aesthetic sense in the act of
writing. If we can get that feeling in our
common usage with computers, we will
have an extremely successful interface."
lamascope developed in Japan
Fels's interest in embodied systems came
to fruition in Japan. After working on a
gesture recognition system for a California
company, Fels took an appointment as a
visiting researcher at Advanced Telecommunications Research (atr) in Kyoto. It was at
ATR while working on an artificial intelligence project that Fels began development
of the lamascope.
The lamascope combines video, audio
and gesture-recognition components that
translate the gestural input of a user into a
shifting kaleidoscope of music and imagery.
continued on backpage
if are €&n make interface.'- that f^i mad, (hat appeal ta
vut emvtivTutl sefi:it>ili^- dtriire td Hit theti itriiljailon:
Spring 99 Exploring the boundaries of complexity
Nick Pippenger unravels
problems in complexity theory
and computational topology
Nick Pippenger, a CICSR member
and professor of Computer Science, inhabits
a field at the intersection of mathematics
and computer science. He explores computational problems from a mathematical
perspective, and although the research is
abstract, it builds the foundation for future
solutions in day-to-day computing.
"I am more of a problem-solving type of
researcher than a theory-building one," says
Pippenger, a specialist in complexity theory
and computational topology.
Complexity theory and
computational topology
Complexity theory seeks to establish as
precisely as possible the number of basic
components or operations needed to carry
out various tasks. It also plays a role in
determining the difficulty of such problems
as knot recognition and knot equivalence in
computational topology.
"The unifying theme of my work is an
interest in mathematical methodology,"
explains Pippenger, who has an engineering
background. "If I can't formulate a problem
mathematically, then I won't work on it."
One problem that has occupied
Pippenger is determining whether or
not a closed curve embedded in space is
unknotted. A closed curve without self-
intersections in three-dimensional space is
said to be unknotted if it can be deformed
continuously, without breaking it or passing
it through itself, to lie in a plane. Until
recently, no good algorithm for recognizing
unknottedness was known.
Knots and polymers
The question has intrinsic significance to
topologists, but has importance as well to
polymer physicists looking at the entanglement, or knotting, of polymers. For example,
scientists examining dna—a biological
polymer—are using mathematical models to
analyze how certain enzymes allow dna
strands to pass through each other.
"A great deal of effort has been devoted
to a quest for algorithms for recognizing
unknottedness," says Pippenger. "It was
only in the 1950s that it was first proved that
this question could even be answered using
a computer."
First bounds on problem
As Pippenger was working on a paper
giving the first bounds on the complexity of
recognizing unknottedness, he learned of
two other researchers working on the
problem. Pippenger and Joel Hass, of the
University of California at Davis, and
Jeffery Lagarias, of AT&T Labs in New Jersey,
produced a paper together which they
presented at the ieee's 38th Annual Symposium on Foundations of Computer Science.
"The knot is one of the more fundamental objects in topology," says Pippenger.
"There is a great deal of intrinsic mathematical interest in the questions involved."
Pippenger's expertise in knots also led
him recently to serve on the supervisory
committee for CS graduate student Rob
Scharein's PhD thesis. Scharein has created
a software program called KnotPlot that
generates complex knot images.
Pippenger next wants to determine the
complexity of recognizing knot equivalence.
Exploring mathematical richness
"I'm interested in exploring the richness
of mathematical situations," says Pippenger,
summing up his approach. "Because of the
nature of certain questions, definitive
answers cannot be obtained by experimental
or empirical methods. Progress is made
exclusively through mathematical methods."
Such methods may seem abstract but as
Pippenger notes, "The transformations in
physics which have occurred in this century
have shown us embodiments in the physical
world of things that were previously known
only in people's minds."
For more information,contact Nick
Pippenger at (604) 822-4030,or
FOCUS Real-time systems: multimedia delivery
and landmine detection
Mabo Ito works to increase
Multimedia and landmines have little
in common, but to Mabo Ito they pose a
similar challenge in computer and software
engineering. CICSR member Ito wants to
improve the performance of real-time
multimedia delivery over Internet Protocol
(ip) networks and the detection of
landmines in battlefield conditions.
Real-time systems require fast data
processing, and accurate error detection and
correction methods, to ensure acceptable
system performance.
"In a real-time system there is often no
time to retransmit the data—you have to
work with what you get," says Ito, a professor in ubc's department of Electrical and
Computer Engineering.
Real-time challenges
Traffic on the internet, and the demand
for video, is growing fast. Real-time encoded
video faces two problems: congestion, which
in turn leads to dataloss, and delay. During
high-traffic periods, routers (that switch and
direct data packets over the internet)
commonly lose packets.
This does not present a problem with
e-mail messages and static data files because
packets can be resent. But this solution fails
with real-time multimedia and video
transmission because they require a continuous stream of live data. When packets are
lost or delayed excessively, image quality
suffers with gaps, bands and other artifacts
appearing in the video image.
"What we've been doing is trying to
understand what kinds of errors occur and
their severity," says Ito. "We've found that
the objective measurement of quality does
not correlate with a subjective view of
quality. Small errors are more visible and
annoying in relatively static video images,
such as a news broadcast, and less noticeable
in high-activity scenes."
Armed with an understanding of these
errors, Ito is developing correction algorithms which can fix errors on the fly. For
example, these algorithms can detect badly
affected areas in a video frame and restore
them by using adjacent areas of the image to
fill in the degraded area. Error-riddled
frames can also be totally replaced by
neighbouring frames.
Error prevention at the source
These techniques work at the receiver's
end; Ito is also looking at error-prevention
at the source. Such methods include error
correction codes, transmission of multiple
picture headers, and prioritizing video
transmission. This last approach—known as
differential service—flags certain kinds of
encoded video packets and gives them a
higher priority in internet traffic.
Ito has been working on this project with
Gerald Neufeld of Computer Science (also a
CICSR member) for the past two years.
Neufeld (currently on leave in Silicon
Valley) has been looking at the protocol and
network aspects of the problem.
The project is part of a major undertaking in network resource management by the
Canadian Institute for Telecommunications
Research (citr) , one of the Networks of
Centres of Excellence administered by
real-time system performance
nserc. Hewlett Packard Canada has
provided strong support, including funding,
equipment and advice. Nortel has also
contributed funding.
Landmine detection
Landmines became an area of interest
when one of Ito's graduate students went to
work at the Canadian Forces' Defence
Research Establishment in Suffield, Alberta.
A research project undertaken there to find
buried, unexploded shells using magnetic
pulse induction led to Ito's current work in
landmine detection.
Ito developed a real-time system to detect
mines at or near ground surface using an
airborne active infrared system. Infrared
lasers scan the ground, and sensors analyze
the resulting light reflection and absorption
patterns. Since all materials—including
plants, earth, metal, plastic, and paints—
have a characteristic reflection signature, the
system can scan a cluttered field and detect a
landmine according to its signature.
A further development of this system
relies on a passive, airborne multispectral
scanner to measure heat and radiation
signatures. Again, different materials have
characteristic reflective properties, and these
signatures are used to distinguish objects on
a battlefield, ^fet another development is the
use of a passive, vehicle-mounted, forward-
looking infrared sensor to provide real-time
detection of landmines.
"The last thing you want to do on a
battlefield is to stop, even in defensive or
peacekeeping situations," says Ito. "This
system can tell you whether there are mines
there or not, and help identify a safe route."
Ito's detection system is a timely development as the global ban on landmines comes
into effect. He also has a number of other
interests, including requirements
specification and system implementation
using programmable logic controllers.
For more information, contact Mabo Ito at
mito@ece.ubc.ca or at (604) 822-4572.
Spring 99 How the Giraffe got its spots
Alain Fournier seeks better ways to simulate and animate natural phenomena
Computer animation has quickly grown in
power and sophistication. Witness the
change in the world of entertainment, where
computer graphics have migrated from
video games to the big screen and taken a
leading role in such Hollywood blockbusters
as Toy Story.
Computer animation continues to
develop as programmers seek better ways of
creating realistic effects. Surfaces and
textures—illuminated and shadowed by
falling light—present a number of problems, which CICSR member Alain Fournier
wants to solve.
Rendering natural phenomena
Fournier, a professor of Computer
Science, specializes in the computer modelling and rendering of natural phenomena.
In earlier work, he built a simple but
realistic model to simulate ocean waves, and
used fractals to create terrain. Now Fournier
is focussing on the generation of natural
textures and surfaces, such as the coats of
patterned mammals like the giraffe and
Working with graduate student Marcelo
Walter and postdoctoral fellow Daniel
Meneveaux, Fournier has created a pattern
generation system that automatically
integrates mammalian coat patterning with
body growth and animation.
Normally, surfaces such as a giraffe's
reticulated coat are generated separately
from the modelling of the animal's geometry. The patterned surface and geometry
are then integrated—texture mapped—to
render the animated animal.
"We are pretty good at getting an
animal's geometric model," says Fournier,
"but it is harder to map a realistic pattern to
its surface without distortion."
In the new pattern generation system,
the pattern is "grown" in tandem with the
animal's geometry. Fournier and his
partners have achieved this by simulating
the biological process of pattern growth,
according to the clonal
mosaic (CM) model. The
CM model accurately
describes natural growth
processes on the skin of
mammals from embryo
to adult. In the case of
the giraffe, it describes the
spatial arrangement and
growth of pigment-
producing cells in the
giraffe's skin as the animal
Fournier and his team
apply a simplified CM
model to generate
computed "cells." These
cells are tied to the
geometric model by
"growing" them directly
on the polygons that
compose the model. The
model itself is described
and controlled by a
cylindrical coordinate system that covers the
major body parts. With the input of real
giraffe measurements, a model can be grown
from embryo to adulthood with a realistic,
corresponding patterned coat. And that
model can then be animated.
Integrating growth & patterning
"The point of integrating growth and
pattern processes is to create more accurate
models of patterned animals," says Fournier.
The pattern generation system is attractive because it can produce a large number
of animal patterns with a relatively small
number of parameters, which can be applied
to a variety of shapes. As well, the cylindrical
model can generate unique, individual
giraffes by inputting real giraffe measurements.
Fournier is also interested in computer-
simulated illumination. He is currently
exploring how to accurately reproduce the
reflective properties of woven fabrics such as
.HI       I) 1
velvet and silk. These fabrics, along with fur
and hair, share self-shadowing properties
that are hard to simulate.
"Self-shadowing is the effect of the
material casting a shadow on itself," says
Fournier, pointing to silk as an example. "It
is an important visual property that is part
and parcel of the texture."
These effects, however, are difficult to
compute, so Fournier is examining different
filtering techniques that simulate the self-
shadowing look.
Technology increasing in sophistication
"As the technology increases in sophistication, people's expectations increase as
well," says Fournier. "So our simulations
have to keep up with the technology by
being more and more detailed. It's a great
time to be in computer graphics!"
Alain Fournier's e-mail is fournier@cs.ubc.ca,
or call (604) 822-6770. Check out his website
at www.cs.ubc.ca/spider/fournier/home.html
FOCUS CICSR Passing Notes
New Master of Software Systems
A new CICSR Masters program will help
fill the growing number of jobs in the
information technology sector. The Master
of Software Systems (MSS) is designed for
students with Bachelor degrees in areas such
as the mathematical and physical sciences,
operations research and engineering (other
than computer science and computer
engineering). The MSS is a 30-credit
program over 3 semesters and will take 16
months to complete. Students must have
some computer program design and data
structures knowledge. Students who do not
have the appropriate course requirements,
but have the necessary technical experience
or background, may be accepted for the
program. The MSS is scheduled to begin in
September 1999, subject to approval by the
BC Ministry of Advanced Education,
Training and Technology. The program is a
joint effort of the departments of Computer
Science and Electrical and Computer
Engineering. For more information see
www.cicsr.ubc.ca/mss .html
CICSR members elected as IEEE Fellows
CICSR Director
Rabab Ward and
ece professor Guy
Dumont were
recently elected
Fellows of IEEE.
Their election
acknowledges their outstanding contributions to the field of electrical engineering.
The IEEE cited Ward's work in digital signal
processing and Dumont's contributions to
the process industries.
Isaacson appointed EIC Fellow
Dr. Michael Isaacson, Dean of Applied
Science and a professor of Civil Engineering, has been elected a Fellow of the
Engineering Institute of Canada.
O'Dor & Croft
World Automation Congress award
Matthew O'Dor,
a recent Masters
graduate in Mechanical Engineering,
received the Best
Student Paper Award
at the 1998 World
Automation Congress (wac). The paper, co-
written with his supervisor Elizabeth Croft,
is entitled "Identifying Salmon Can-Filling
Defects Using Machine Vision." wac is the
major international symposium devoted to
automation; it was held in May 1998 in
Anchorage, Alaska.
New Networks of Centres of Excellence
CICSR faculty members will participate in
two of the three new Networks of Centres of
Excellence (nce) recently announced by the
federal government. David Kirkpatrick,
Jack Snoeyink and Craig Boutilier (cs) will
take part in the Mathematics of Information
Technology and Complex Systems Network
(mitacs); Kirkpatrick and Snoeyink will also
join the Geomatics for Informed Decisions
Network (geoid). The new nces, including
the Canadian Arthritis Network, will share
$4i-million over the next four years.
Business/Education Partnership award
Michael Jackson (ece), his graduate
students and Thomas & Betts Photon
Systems Inc., have won the first Business/
Education Partnership Award from the BC
Science Council. The award was given for
their collaboration in fibre optics technology.
Tan back from teaching in Mexico
Joseph Tan (a CICSR associate from the
Faculty of Medicine) has returned from a
special teaching engagement on behalf of
CICSR at cetys in Mexico. Tan co-taught a
course in Analysis of Decision Processes to
students enrolled in the university's DEng
(Doctoral in Engineering) program from
September to December 1998.
Killam prizes for
Bond, Kallel &
Greg Bond (ece),
Samir Kallel (ece),
and Jack Snoeyink
(cs) are among the
most recent crop of
Killam award
winners. Bond won a
Killam Teaching
Award; Snoeyink was
awarded a Killam
Research Prize; and
Kallel won an Isaac
Walton Killam
Memorial Fellowship.
Jack Snoeyink
Vinod Modi wins AIAA award
The American
Institute of Aeronautics and Astronautics
(aiaa) has awarded
Vinod Modi (me) the
1999 aiaa Pendray
Aerospace Literature
Award. The award,
the aiaa's highest,
honours Modi's extraordinarily significant
contributions to the literature of aerospace
vehicle dynamics, controls and robotics, as
well as his teaching.
Challenge CICSR students
to solve them!
Does your company or organization
have a problem that needs fixing?
Why not challenge a CICSR student to
develop a solution! We invite you to
post a $500 award, and the CICSR Office
will match an appropriate student to
your particular problem. If you are
interested, please contact Gale Ross in
the CICSR Office at (604) 822-6601 or at
ross@cicsr.ubc.ca for more details.
Interactive Video-on-Demand Systems:
Resource Management & Scheduling
ed. Babak Hamidzadeh
and T.P Jimmy To
(Boston: Kluwer Academic
Publishers, 1998)
Interactive Video-on-Demand Systems
addresses issues in scheduling and management of resources in an interactive
continuous-media (e.g., video and audio)
server. The book emphasizes dynamic and
run-time strategies for resource scheduling
and management. Such strategies provide
effective tools for supporting interactivity
with online users who require the system to
be responsive in serving their requests, and
whose needs and actions vary frequently
over time.
Intuitive interfaces, continued from page 3
The device has drawn international attention—last year, the lamascope won first
prize at the Petrobas Virtual Reality exhibition in Brazil and was highlighted at Opera
Totale in Venice. It is now on exhibit at the
Museum of the Future in Linz, Austria.
Fels has noted with interest the active
and passive interactions of people using the
lamascope, where control shifts from the
user to the machine and back again. The
relationship of user to machine is not just a
physical interaction, but a cognitive and
emotional one as well.
Computational Intelligence:
A Logical Approach
David Poole, Alan
Mackworth & Randy
Goebel (New York: Oxford
University Press, 1998)
Computational Intelligence: A Logical
Approach provides an integrated introduction to artificial intelligence (ai). It weaves a
unifying theme—an intelligent agent acting
in its environment—through the core issues
of AI. The authors develop AI representation
schemes and describe their uses for diverse
applications, from autonomous robots to
diagnostic assistants to infobots that find
information in rich sources. For more
information about the book see
Health Decision Support Systems
■talk 1*1 —
'-MP<i hum
ed. Joseph Tan with Samuel
Sheps (Gaithersburg: Aspen
Publishing Inc., 1998)
Health Decision Support Systems covers the
theories, methods and applications of
decision support technology in health care.
It explores how medical and health care
decision making can be effectively supported
through linked databases, simulated models,
and intelligent graphical user interfaces. The
book covers the use of data mining techniques, neural networks and other expert
methods. The book also illustrates how
health decision support system technology is
applied in the health care field.
"How can we understand these interactions to make better interfaces?" Fels asks.
"If you can develop an interface that people
can bond with, they're going to be able to
be expressive with it—even if it's an on/off
switch. Whatever it is, it's going to be
satisfying. I'm working on developing that."
Fels's intelligent agent and intelligent
environment projects also focus on designing with the emotional, cognitive and
physical constraints of the user in mind.
Currently, Fels is extending the Glove
Talk II system into the musical domain. He
is in the process of developing virtual
musical instruments which can be played
Fels's other projects and interests include
using neural networks to perform system
identification and non-linear data compression with constraints. His work on the
lamascope was funded by ATR MIC Research
Laboratories (www.mic.atr.co.jp), while
Glove Talk II received funding from nserc
and iris.
For more information contact Sid Fels at
(604) 822-5338 or at ssfels@ece.ubc.ca
CICSR Centre for Integrated Computer Systems Research www.cicsr.ubc.ca
The UBC Centre for Integrated Computer Systems Research (CICSR) is an interdepartmental
research organization made up of computer-related research faculty members in the
departments of Computer Science, Electrical and Computer Engineering, and Mechanical
Engineering. Currently, there are more than 70 CICSR researchers who direct over 300
graduate students and collaborate with dozens of industrial firms in areas such as robotics,
artificial intelligence,communications,VLSI design, multimedia, and industrial automation.
Return Address:
CICSR, University of British Columbia
289-2366 Main Mall,Vancouver, BC,V6T 1Z4
Editors:  William Knight, Linda Sewell
Design:   wilyum creative
Photos:  Janis Franklin,
Biomedical Communications
Office:   University of British Columbia
289-2366 Main Mall
Vancouver, BC, Canada, V6T 1Z4
Tel:   (604)822-6894 Fax:(604)822-9013
E-mail:   cicsrinfo@cicsr.ubc.ca
Contact: Linda Sewell, Publications Coordinator,
CICSR Office


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items