This article from NYTimes.com
has been sent to you by carter@sncarter.com.
Regarding the recent discussions about input methods, this article implies that within the near future (I'd guess 5 years) the keypad will be gone. I've included the whole text of the article, since the NYT is one of those sites you have to register for.
Steve Carter
/-------------------- advertisement -----------------------\
<A HREF="http://ads.nytimes.com/RealMedia/ads/click_lx.ads/web59t.lga2.nytimes.com/RealMedia/ad.html/0/Right2/default/empty.gif/64323134306634303361356166346430" target="_top"><IMG SRC="http://ads.nytimes.com/RealMedia/ads/Creatives/default/empty.gif" BORDER=0 WIDTH=2 HEIGHT=2 ALT="" BORDER=0></A>
\----------------------------------------------------------/
Controls Rely on the Twitch of a Muscle, Not the Twitch of a Mouse
February 8, 2001
WHAT'S NEXT
By CATHERINE GREENMAN
COMPUTERS, cell phones and other gadgets continue to shrink, but the
means of interacting with them, like keyboards and keypads, are
either too large and cumbersome or too small for the human hand.
That is the belief of a group of NASA scientists who have been
working on more direct ways to control machines, starting with jets
and spacecraft.
Speech recognition is one alternative, but it is not ideal for
cockpits or other noisy places. So the scientists are attempting to
show how gestures can link humans more directly to the devices they
control. They recently demonstrated that electronic signals picked
up from human muscles could be used to land a jetliner in a
simulation. These kinds of controls could someday be used in things
like video games as well.
"What's driving this project is the belief that we've become
increasingly disassociated from the actual state of our physical
environment," said Dr. Charles Jorgensen, the lead scientist on the
project. "We're looking at whether we can find a way of more
tightly integrating ourselves with the machines around us." Dr.
Jorgensen is head of the neuroengineering laboratory at the space
agency's Ames Research Center in Moffett Field, Calif.
To demonstrate gesture-based, or bioelectric, control, Dr.
Jorgensen and his team fashioned a five-inch-wide armband from
exercise tights, and sewed on eight metallic buttons to function as
dry electrodes, receiving electrical signals from muscles in the
pilot's arm. Wires carry signals from the electrodes to a computer.
The NASA team has tested the armband in about 100 trial runs in
the past six months, landing a 757 passenger jet simulator at a
virtual version of San Francisco Airport. The simulation software
ran on two personal computers with 17-inch computer screens; one
displayed a simulation of a cockpit control panel, while the other,
a flight simulator, showed the position of the jet from multiple
angles.
Before the trial runs, the scientists designed pattern recognition
software, a variation of neural net software, which can be
programmed to identify repeated physical patterns.
Dr. Kevin Wheeler, a scientist who worked on the project, acted as
the pilot during the trials. Once the software was able to
recognize the electrical signals from the arm as movements, Dr.
Wheeler sat in front of the simulated cockpit at the control
terminal, closing his fist as if he were closing it around an
actual control stick, then moving his arm as if he were shifting
the control stick into banking and pitching motions. As he moved,
the electrical signals in his arm were picked up by the button
electrodes on the armband and transmitted to a computer a few feet
away through a wire. Pattern recognition software then digitized
the signals and recognized the patterns as gestures; signals keyed
to each gesture were then transmitted over a network to the flight
simulator on the other terminal.
Dr. Jorgensen said controls based on the electrical signals from
muscle movements would allow pilots to connect more directly and
intuitively with aircraft and spaceships, without joysticks or
other controls. Flying a plane with gestures instead of manual
controls, he said, would also cut costs.
"A lot of the cost differentials in the building of jets are made
up of having a separate compartment in them specifically designed
to wrap around a pilot in a particular location in the aircraft,"
he said. "There's big design savings by taking away that need."
Eliminating conventional control devices in a space vehicle would
have several benefits, Dr. Jorgensen said. It would certainly help
an astronaut wearing a spacesuit with bulky gloves, which would
make it difficult to type and manipulate controls. An armband with
electrodes that detected the astronaut's muscle movements inside
the spacesuit and transmitted commands would let the astronaut
carry out these functions more easily.
"If there's an accident and there's depressurization," Dr. Wheeler
said, "you can't get out of a spacesuit. If the computers were
still functional, you'd still have limited means to interact with
them."
In the event that an astronaut sends out a robot on a space walk,
the astronaut could use his own arm movements to control the robot,
instead of pushing buttons or using other manual controls. The
robot would mimic the astronaut's own motions.
Dr. Jorgensen said that NASA had applied for a patent on the
pattern recognition method and bioelectric armband concept and that
he hoped to see the first commercial applications for the
technology within a couple of years. The first commercial products
would probably be computer game controllers, which do not require
the same level of precision as a virtual aircraft controller, he
said. Virtual keyboards and mice would be the next logical step.
"We're envisioning something that looks like a wristwatch on your
arm," Dr. Jorgensen said, "which would enable someone to type on a
photocopy of a keyboard that could be thrown away." The virtual
keyboard could also be a blank, flat surface, assuming that someone
is a good typist and does not need to see the keys.
Further down the road, it may also be possible to use bioelectric
signals to produce "silent speech." The scientists are exploring
ways in which electrodes could transmit signals from muscle
movements in the tongue as words were formed within the mouth — but
not actually spoken — to a computer that could translate the
signals into digitized speech.
"The idea would be to get away from having to hear people talking
on their cell phones," Dr. Jorgensen said. "You could more or less
sit there and communicate and have it be translated
electronically."
For now, the main challenge is honing the system to recognize the
subtle arm and wrist turns that pilots typically make during a
flight. The group is also working on ways to make the software
recognize and compensate for different arm sizes and for the
different ways individuals move.
"We obviously have a ways to go yet," Dr. Jorgensen said, "and
haven't licked all these things. The analogy would be with speech
recognition software. It works pretty well out of the box, but you
need to custom- train it for a while."
http://www.nytimes.com/2001/02/08/technology/08NEXT.html?pagewanted=2?ex=982644501&ei=1&en=811451873318cde9
/-----------------------------------------------------------------\
<A HREF="http://ads.nytimes.com/RealMedia/ads/click_lx.ads/web59t.lga2.nytimes.com/RealMedia/promo.html/0/Right2/default/empty.gif/64323134306634303361356166346430" target="_top"><IMG SRC="http://ads.nytimes.com/RealMedia/ads/Creatives/default/empty.gif" BORDER=0 WIDTH=2 HEIGHT=2 ALT="" BORDER=0></A>
\-----------------------------------------------------------------/
HOW TO ADVERTISE
---------------------------------
For information on advertising in e-mail newsletters
or other creative advertising opportunities with The
New York Times on the Web, please contact Alyson
Racer at alyson@nytimes.com or visit our online media
kit at http://www.nytimes.com/adinfo
For general information about NYTimes.com, write to
help@nytimes.com.
Copyright 2001 The New York Times Company
[ Did you check the archives? http://www.appelsiini.net/keitai-l/ ]
Received on Thu Feb 8 16:52:42 2001
