Graduate student Tovi
Grossman, part of the human-computer interaction
research group at U of T, is figuring out how we
all will be using computers in 20 years.
Lots of
people have great ideas about new ways of
interacting with computers. But building a new
device is only the first step. The mouse, for
example, was invented in the 1950s, but it took
decades before everyone recognized how it could be
exploited productively. Grossman’s job is to
investigate prototype computer devices and
discover how we might want to actually use
them.
Natural curves
This
summer, before starting course work in his
master’s program, Grossman finished a project with
a device called a “shape tape.” It looks like a
metre-long band of blue plastic, and inside the
band are optical fibres and sensors that can
detect bending and twisting. With one end of the
band fixed in place as a reference point, it can
provide digital information about its own position
in three dimensions.
Although shape tape
existed before Grossman laid his hands on it,
there was no clear way of using it productively.
His summer project was to create software that
uses the input from shape tape to draw and edit
3-D shapes in a computer. Grossman developed a set
of simple gestures that anyone can learn and use
to activate different computer functions. For
instance, holding the shape tape in both hands and
swiftly moving them downwards and together, as
though breaking a stick, could tell the computer
you want to switch from drawing curves to drawing
angles.
Grossman wrote 9,100 lines of
computer code during the summer to create the new
software for the shape tape. But that only proved
that a new interface could be built, not that it
was useful. To see whether anyone would actually
want to use the device, he invited several graphic
designers into the lab to give it a try.
In
a series of trials, designers were asked to
perform tasks with the shape tape that they would
normally perform with a mouse or an electronic
stylus. They reported that although shape tape
would not replace their normal input tools for
most tasks, they did find it was a great advantage
whenever they wanted to refine the shape of a
curve.
Normally, editing a curve in the
computer means using the mouse to change the
position of dozens of inflection points. But with
shape tape in hand, the designer just nudges or
stretches the curve into its new formation. That
saves a lot of time and is a more natural way to
manipulate curves.
Grossman agrees that
shape tape is not a replacement for the mouse, but
argues that it will find a role. In a few years,
he predicts, “shape tape will be one tool in every
graphic designer’s tool kit,” ready to use
whenever a natural-looking curve is required.
Research is fun
The
latest excitement at the human-computer
interaction lab is the recently-delivered
prototype of a “volumetric display,” built by
Actuality Systems. In an amazing improvement over
the simulated 3-D that can be had on typical
computer screens, the machine can display real
three-dimensional images within its transparent
sphere. As you walk around the sphere, you see all
sides of an object projected inside.
Inside
the clear dome is a screen that spins very
rapidly, onto which an image is projected from
below. As the screen spins, the projected image
varies perspective in such a way that from any
position, we perceive a volume-filling object
hovering under the dome. The image may be small
and fuzzy—but just think about the quality of an
early computer screen compared to one of today to
imagine the difference that a few decades can
make.
Grossman demonstrates the machine by
pulling up a model of the sucrose molecule, then a
human skeleton in full detail, then a crude
helicopter that can fly over a wire-frame mountain
range. “You can’t crash the helicopter,” he
assures. “Everybody tries.”
The display
is very cool. But is it useful?
The
possibilities for 3-D modelling and for
visualizing complex data with such a machine have
barely been glimpsed. How can the machine’s
abilities be exploited by scientists? By
designers? By artists? If we can simulate three
dimensions on a flat monitor, can we simulate four
inside a volumetric display? When you look at the
display, the first thing you want to do is touch
the image. What kind of input devices can we use
to make that impulse practical?
Grossman is
one of the first researchers to use and program a
working volumetric display, so he will be able to
start answering some of those questions. The goal
for his master’s program is to demonstrate the
potential of volumetric displays and convince
everyone that this is the way 3-D will be done 20
years from now.
Does it seem like his job
is to play with expensive toys? Grossman agrees,
grinning. “Research is fun,” he
says.
