Technological advances in the field of human-computer interaction
will pay off dividends for U.S. military programs focusing on battlefield
robots and tactical data networks, experts said.
Jim Osborne, a robotics specialist with the Pittsburgh Robotics
Initiative, believes that autonomous machines today have enough
intelligence to govern their own actions.
But not everyone agrees. Some experts believe that autonomous capabilities
are mature and advanced enough for full deployment in robotic platforms,
while others advocate a more cautious approach.
Osborne predicts that U.S. military programs will support a “mixed
mode of operations,” including both fully autonomous robots
and others that need more human control.
The “program to watch” for robotics technology development
is the Army’s Future Combat System, said Osborne. Under this
project, the Army plans to develop a fleet of light combat vehicles,
some of which may be remotely operated. “The success of this
program will shape the future of military robotics,” Osborne
said. “We’ll see how well the rubber meets the road.”
Additionally, said Osborne, there are many other areas where the
U.S. military services will benefit from robotics technology. These
include de-mining, and search-and-rescue missions.
Some of the challenges in developing advanced robots can be attributed
to the fact that “computers are hard to use,” said Clinton
Kelly, senior vice president of advanced technology programs at
Science Applications International Corporation, in San Diego. “It
is unarguable that computers have changed the way we do business,
but it is debatable whether or not they have made it more productive,”
Kelly told a conference of the Government Electronics and Information
Technology Association (GEIA).
Computers make demands on certain cognitive abilities such as logical
reasoning and spatial memory, he explained. Users who score in the
top 25 percent in logical and spatial ability do twice as well with
a computer as the lowest 25 percent. “We concluded that one
out of three college-educated people can’t use computers very
effectively,” Kelly said.
More research work, therefore, is needed in human-computer interaction,
he stressed. One of the most promising areas is speech recognition.
Systems are available today for about $200. The software available
has a large vocabulary, from 50,000 to 100,000 words in multiple
languages. The training time has been reduced to about three to
10 minutes, with a 98 percent recognition accuracy, Kelly said.
One growing application for voice recognition technology is to retrieve
one’s e-mail.
Interface devices will lead to what Kelly calls “the age
of proxy devices” or “the post-PC era” in the
next three years. He believes that general-purpose computers will
be replaced with simpler devices, set up for specific tasks such
as e-mail or Web-surfing.
While speech recognition—or converting audible signals to
digital symbols is a tough problem—an even harder one is getting
computers to understand natural language, the actual meaning of
words. Kelly summed up the problem of semantic ambiguity. “You
take a word like ‘strike’ and it has something like
20 to 40 definitions. The average common noun out of the top 200
in usage has about eight meanings and the average common verb about
12 meanings.”
There are syntactical ambiguities as well. Any one sentence could
potentially have at least half a dozen meanings, Kelly explained.
The way to deal with these problems is to develop computers that
know and reason, according to Kelly. “Turns out we’ve
been doing that in the machine intelligence community for a long
time. We have created systems that are called knowledge-based systems
or expert systems.” Kelly said. He cited a project known as
Cyc, short for encyclopedia, at Cycorp in Austin, Texas. “They
have something like a billion actions in their database that reflects
about 10 years of work. ... The kind of knowledge people need to
do anything has to be understood by the machine in order for it
to understand natural language.”
By 2020, Kelly believes, “the age of thinking machines will
be well underway.”
Computers in Combat
As the military services try to take advantage of computer capabilities
in combat applications, they should focus on interface devices,
said Corinna E. Lathan, chief executive officer of AnthroTronix
Inc., in College Park, Md.
The company is involved in two information technology and robotics
programs currently funded by the U.S. Defense Advanced Research
Projects Agency (DARPA).
One is the digital military police (MP) program, managed by the
U.S. Army Soldier’s System Center. Digital MP is a wearable
communication and information management computer. The system was
developed by ViA Inc., MicroOptical Corporation and Honeywell Inc.
A pair of eyeglass frames contains a built-in miniature camera—used
for face recognition and image displays. A noise-canceling microphone
and earphone are used for voice recognition.
The camera allows streaming video to be transmitted from one MP
to another. Video generated at checkpoints can be matched against
mug shots from the National Crime Interdiction Center database,
for example. A military “e-Book” can be used to plot
maps that can be shared between soldiers. The e-Book display gives
off no light, and the display is readable in sunlight or starlight
with night vision goggles. There is also an electronic glove, which
works as a computer mouse for the e-Book. It can capture combat
gestures and communicate the meaning to other troops in the area,
even those not in the line of sight. For instance, a soldier can
hold up a fist in a gesture that means halt. His wearable computer
translates that into the verbal command and sends it to the earphones
of other soldiers in the area. The gestures are preprogrammed into
the various sensors contained in the glove.
AnthroTronix originally used this technology in systems developed
for children with cerebral palsy, who do not have the fine motor
skills to work with objects such as keyboards.
Lathan cited the example of one boy who loves sports. With a baseball
glove—equipped with an accelerometer—his movements can
be recorded. The computer can then be programmed to accept different
gestures as commands. If he wants to swing a bat in an electronic
baseball game, all he might have to do is raise his arm, instead
of using a joystick or control pad. This can also help him increase
his range of motion so that he possibly could feed himself, Lathan
explained.
The digital MP also can be programmed to translate English to Spanish,
Korean, Arabic, German, French, Italian, Portuguese, Dutch, Thai
and Turkish, with only a five-second delay. There are plans to add
a feature that allows military terms such as “clicks”
to be translated into “kilometers.”
Another project that is pushing the technology in computer interface
is the tactical mobile robot (TMR). The program seeks to develop
lightweight remote-controlled vehicles that can perform reconnaissance
missions and other tasks in areas considered too dangerous for manned
patrols.
In the early stages of the program, the robot was controlled with
a laptop computer, which was not “very effective,” according
to Lathan.
The obstacles robots used in ground operations face are very complex,
explained Army Lt. Col. John Blitch, the program manager for DARPA.
They not only have to deal with obstacle detection and avoidance,
but with negotiation as well. “These robots have to climb
and maneuver on obstacles,” Blitch said in a recent interview.
“They might have to cut through a chain link fence.”
The NASA Jet Propulsion Laboratory has also created a robot that
can climb stairs. This robot possibly could interact with hostile
or non-hostile humans, according to Blitch. Even dogs or horses
have been known to interfere with the robots.
While ideally Blitch would like to see the TMR gain full autonomy,
dealing with these kinds of situations requires a strong human interface.
“We’re talking about a robot supervisor, not a driver,”
stated Blitch. “The robots are still pretty dumb. We need
to take over sometimes, override internal commands every once in
a while.”
Everything used to control a TMR is wearable and part of a soldier’s
normal combat gear, explained Blitch. The primary interface is a
pair of gloves that act as regular protection until a button is
pressed. They then become gesture recognizers that control the robot.
The thumb acts as a joystick. Touching the pads of various fingers
together or rotating the user’s wrist will send different
commands to the robot. Sign language is also used instead of speech
recognition, which can be misinterpreted by the robot or detected
by enemy forces, Blitch added.
The robot is equipped with an Omni-Cam, a camera aimed at a hemispheric
mirror allowing a 360 degree view. The image is sent to an overhead
display set in a pair of glasses worn by the supervisor.
Auditory information is communicated through a bone phone, a tight
connection to the skull. “Sound resonates through bone,”
explained Blitch. “The sound is sent through the skull to
the inner ear to the brain.”
The robots currently need a lot of supervision, conceded Blitch.
This is why he advocates a new operational specialty for robot operations.
Studies show that control of the robot can degrade a soldier’s
primary skills.
Gaurav Sukhatme, a professor of computer science at the University
of Southern California (USC), described some of the work that the
university is doing on the TMR project. “We have developed
a touch-screen-based interface for commanding and controlling a
group of TMRs in support of complex missions,” he said. The
user interface work began in-house at USC and has since been subcontracted
to two companies: Rossum Technologies and Indelible Systems.
USC also is participating in the development of autonomous aerial
vehicle control and cooperative indoor mapping by robot groups,
explained Sukhatme. “In the aerial vehicle-control area, we
have demonstrated how an onboard computer can be used to control
a helicopter robot, so that a war-fighter (who is not a helicopter
pilot) could task the flying robot at a high level (point and click)
to performs missions, such as surveillance or deploying other robots.
“In the cooperative mapping area, we have demonstrated how
a group of small mobile robots can autonomously create a map of
the inside of a building. These robots are not joysticked by a human.
Instead, they search and map out the building on their own. A map
of the inside is displayed to a human on a laptop.”