Pearl flirts with old men. But women like
her too. Her wide eyes and smiling lips lend her face an expression
that hovers between vacuous gaiety and humble servility. But don’t
let the vacant look fool you. Pearl is one sharp cookie.
She whizzes about reminding her clients
to eat, drink, take medicine, or use the bathroom. She also guides
the old folks from room to room as she chats about the weather or
TV listings. This sweetheart of the elderly is a self-directed mobile
robot with advanced artificial intelligence to assist people with
the activities of daily living.
She may never win a beauty contest, yet
she’s won the hearts of elderly folks at the Longwood Retirement
Community in Oakmont, Pa., where field testing on the 4-and-a-half-foot-tall
robot nurse is underway this summer.
Her designers say they took special care
to make her pretty, but it’s what’s inside that counts: Two Intel®
Pentium® 4 processor-based PCs run software to endow her with wit
and ability to navigate; a differential drive system propels her;
Wi-Fi helps her communicate as she rolls along; laser range finders,
stereo camera systems, and sonar sensors guide her around obstructions;
microphones help her recognize words; speakers enable others to
hear her synthesized speech; an actuated head unit swivels in lifelike
animation. All this and more is housed in her slender frame.
"We're getting along beautifully," says
one older gentleman as Pearl leads him to a room for his therapy
session at Longwood. "But, I won't say whether she's my kind of
girl," he quips.
Part of an interdisciplinary research initiative focused on
robotic technology for the aged, the nursebot project brings together
researchers from four schools: the University of Michigan, Pittsburgh,
Carnegie Mellon, and Stanford. Funding for the project comes from
several sources, including Intel.
Dollars from the Intel Research Council’s
Applications Committee help University of Michigan’s Martha Pollack
and Satinder Singh develop the robot’s software system. Pollack
is a professor of electrical engineering and computer science, and
an Intel principal investigator.
“The grant from Intel has been tremendously
helpful for us,” Pollack tells Circuit News. “It’s giving
us the opportunity to research reinforcement learning techniques.
We’re programming the robot to learn over time what interaction
strategies work best with people.
“Imagine putting this system in someone’s
home and watching what they react to or don’t react to, when they
seem to be happy and when they don’t, and modifying the robot’s
interaction strategy on a very personal basis. This is what Intel
is helping us do.”
Pearl of great price
Don’t look for a nursebot at your pharmacy or health
care supply store any time soon. A consumer version of the computerized
caregiver is still years away.
For one thing, these robots are very expensive.
Pearl is a one-of-a-kind, costing close to $100,000. To ready a
mass-market version of the automaton, researchers must overcome
a rash of other issues as well, such as battery power and the robot’s
inability to navigate steps.
So in the meantime Pollack has installed
her software, dubbed Autominder, on platforms more affordable than
the robot, such as a PDA and a walker. Although not as appealing
as Pearl, a PDA configured with the software costs far less—about
$500 when packaged with in-home sensors. Pollack’s also installed
Autominder on a walker. Equipped with audio-visual displays and
sensors, the walker, too, is less expensive than the robot and guides
the elderly from place to place.
Researchers see great potential in both
the PDA and the walker, but Pearl gets all the press.
Pearl as professional, pal, or pet?
“We didn’t use the robot just
to gather attention. In fact, I was pleasantly surprised by how
much attention it brought,” says Pollack. “From a research perspective
there are some real advantages when using a robot as a platform.”
For example, some of Pollack’s colleagues
say the robot’s human form may provide a kind of social interaction
that’s good for people who are often very isolated.
“But nursebots will never be a substitute
for real nurses,” Pollack’s quick to point out.
“We’re not trying to replace human caregivers,”
she says. “We don’t think we could or would want to. We just want
to supplement because with the upcoming demographic shift there
aren’t going to be enough people to be full-time human caregivers
for older adults,” she says, referring to the global trend of increasing
life spans and the enormous challenge facing the technical community
researching how to sustain independence and preserve quality of
life among older adults.
The cresting age wave
Intel’s Eric Dishman, manager of Intel’s Proactive
Health Research lab, agrees. He says countries around the globe
will soon face the problems associated with a cresting age wave.
“Starting at the end of this decade, the
first wave of baby boomers will retire and the population of elderly
people will swell,” says Dishman. “Countries must somehow increase
the quality of care for growing numbers of people over 65, the most
expensive demographic to care for, while somehow reducing health
Dishman and other researchers believe technology
is the answer to affordably meet the needs of this demographic,
but many challenges still face researchers developing home health
Dishman and Pollack have been instrumental
in addressing these challenges while gathering broad support for
the cause. Both are active members of the Center for Aging Services
Technologies (CAST), an organization that grew out of Intel’s research
and now has members representing over 300 technology and health
care companies and universities.
“The objectives of the nursebot project
really align well with Intel’s goals in home health care,” says
The robot Pearl provides a research platform
to test out a range of ideas for assisting elderly people. Researchers
hope that such autonomous mobile robots will one day live in the
homes chronically ill elderly persons to perform a variety of tasks,
- reminding elderly patients to visit
the bathroom, take medicine, drink, or see the doctor.
- connecting patients with caregivers
through the Internet. The robot is a platform for tele-presence
technology whereby professional caregivers can interact directly
with remote patients, reducing the frequency of doctor visits.
- collecting data and monitoring the well-being
of patients. Emergency conditions, such as heart failure or high
blood sugar levels, can be avoided with systematic data collection.
- manipulating objects around the home
such as the refrigerator, washing machine, or microwave. Researchers
say arthritis is the main reason elderly give up independent living.
- taking over certain social functions.
Many elderly people are forced to live alone, deprived of social
contacts. The nursebot may help shut-ins feel less isolated.
Pollack and other researchers
say the time has come for mobile robots in the home-care sector.
They point to the technology currently available to develop robust,
reliable robots. They also describe the unprecedented need for cost-effective
solutions in the elderly care sector.
If these researchers are
correct, Pearl and her progeny might forever change the way health
care professionals serve the growing ranks of older folks and take
human robot interaction to new heights.