IEEE Spectrum January, 2007 - 38

Winners & losers 2007 ////////////////////////////////////////////////////////////////////////////////////////////

e
biom

dica

tongue
Vision
A fuzzy outlook for
an unpalatable
technology

IMAGINE BEING BLINDfOLDED
and having an array of electrodes
sitting on your tongue. Around your
neck hangs a flat box containing a
microprocessor, and your lips close
around a long cord that connects
to the box and a camera. the cord
dangles halfway down your chest, as
if you're drooling electronics. While
you tentatively direct the camera,
you feel your tongue vibrate at tiny
discrete points that form a circle.
the electronics in your mouth are
telling you that you are facing a
round object. It might be a tennis ball
right in front of you. but then again,
it might be a hot-air balloon a kilo-

meter away. You really can't tell.
the main idea of the brainport
is to help blind people by translating
visual information into tactile cues.
A video feed is reduced to simple
shapes, which are then drawn on the
tongue by activating certain electrodes, each of which applies a small
voltage that lightly tingles the tissue.
As you turn the camera to explore
an area, the electrodes respond with
different patterns of mild zaps to
indicate the shapes of objects in the
camera's field of view. the sensory
experience of the brainport, in visual
terms, is a flat world rendered in
blurry, monochromatic silhouettes.

IEEE Spectrum | January 2007 | NA

such obstacles as a drop-off at a
stairway or a curb, or a glass door.
"I think it's easy to recognize fairly
simple, high-contrast objects on a
black or a white background-that
can be done," says richard Normann,
a bioengineering professor working
on the university of utah's visual
neuroprosthesis project. "but our
world is not that kind of visual world."
the fact that the system can only
supplement other aids raises a logis-

BrainPort
GOAL: to help
people with serious vision loss navigate
by translating visual
data from a camera into
vibrations on the tongue;
also to assist people with
balance disorders by
providing tactile feedback
on changes in the way the
head tilts.
WHy IT'S A LOSEr:
the technology
seems unlikely to provide
enough clear benefits to
counterbalance the cost
and annoyance of wearing
the device.
PLAyEr:
Wicab.
WHErE:
middleton, Wis.
STAff:
6 engineers.
BuDGET:
uS $6 million to
$6.5 million for vision,
funded through National
Institutes of health and
dArpA grants.
MOrE: http://www.
wicab.us

lEFt: JASoN lEE; rIght: WIcAb

loser

critics say most blind people
would find information at such a low
level of detail, transmitted through
the tongue, to be of limited value in
decoding the clutter and chaos of
everyday life. given its estimated
uS $5000 price tag, doubts about
the brainport's utility are compounded by questions of whether
low-vision consumers will prefer it to
other, cheaper assistive technologies.
developed and marketed
by Wicab, a small company in
middleton, Wis., the brainport is an
offshoot of vision-loss research done
by paul bach-y-rita and colleagues
in neighboring madison at the
university of Wisconsin's biomedical engineering department. the
device's cord is an awkward, potentially embarrassing feature that was
summarily rejected by everyone contacted for this article. Even Wicab's
chief executive, robert beckman,
says, "to be commercially feasible,
we need to be in a wireless format,
where the array is mounted on a tray
that is fixed on your upper teeth."
"We've made an array that
mounts that way," he adds. but the
wireless aspect is still to come.
Although those changes would
definitely improve the device's cosmetic appeal, the inherent shortcomings of the sense of touch impose a
low ceiling on the ambitions of any
tactile vision device. Not only is the
brainport limited by the size of its
array, which consists of 100 electrodes in a 10-by-10 grid, it is also
hampered by the simple fact that
the human brain has not evolved
to process large amounts of tactile
information simultaneously.
Even if the tongue and the array
were able to provide the canvas for
a detailed, pointillist masterpiece
and a palette of strong and weak
buzzes were calibrated to mimic
gray scale, the result would present
a serious cognitive challenge. the
patterns would dissolve into nonsense, the signal lost in an onslaught
of zaps. "It's too hard to learn to
integrate all this information-it's
not natural," says hong tan, an
electrical engineering professor at
purdue university, in West lafayette,
Ind., who works on tactile humanmachine interfaces.
A camera-tongue display combination would not be able to convey
enough detail to let a sightless
person get around safely, so a user
would still need to rely on a Seeing
Eye dog or a cane to contend with

www.spectrum.ieee.org

http://www http://www.wicab.us http://www.spectrum.ieee.org

Table of Contents for the Digital Edition of IEEE Spectrum January, 2007