Back to Clippings Index

Elderly receive a helping hand from new robots
April 27, 2000

SAN FRANCISCO -- When we think of the elderly, we don’t generally associate them with robots. But with an aging demographic in both the U.S. and Japan and the rising costs of care, the need for robotic assisted living will continue to grow. Eager to fill this need, roboticists from around the world got together to present their ideas for rehabilitation robots this week at the robotics 17th annual IEEE International Conference on Robotics and Automation in San Francisco.

Professor Steven Dubowsky of MIT explained that the motivations for using robotic aids for elder care are both economic and qualitative. If you can extend a person’s stay in an independent living facility just a little longer before sending them to a nursing home, it will mean a remarkable financial savings and, says Dubowsky, an improvement to their overall quality of life.

But delaying the transition to a full-service nursing home doesn’t mean you can just ignore the needs of your aging loved ones. “Elderly people get lost in an assisted living facility, they get confused. They need physical support for getting from one place to another. They need their health monitored, they need somebody to hold onto their hand and say, ‘You know, you’re running a little bit of a fever today, maybe the nurse should take a look at you.’ They need to know when to take their medication.”

Dubowsky thinks his device, the Personal Aid for Mobility and Health Monitoring (PAMM), will one day be able to provide all of these services to the elderly. A mobile robot designed specifically for use in independent living facilities, the PAMM robot is available as a part of a cane or a walker. Either way, the robotic functionality is largely the same.

The PAMM unit communicates wirelessly, relaying health information of the user (such as temperature and blood pressure) to the central computer system of the facility. Sonar is used to detect obstacles and small dots painted on the ceiling act as a map, letting the PAMM now exactly where it is in the building and what is nearby. Voice prompts let the users know their location and make decisions about where they would like to go. It also provides an easy way to remind users to take their medicine or eat lunch.

Gerry Lacey and Shane MacNamara designed a similar walker robot at the University of Dublin, except that they decided to focus on the frail and visually impaired. There are more traditional solutions, but a guide dog would require too much care and a cane wouldn’t provide enough support for these people. Their robot, called the Personal Adaptive Mobility AID (PAM-AID), provides the full support of a walker and only needs the basic maintenance of recharging each night.

When they started designing the first prototype in 1994, the PAM-AID had the same kind of motorized propulsion system as the PAMM. That idea was abandoned after it became clear that many users simply couldn’t keep up with the robot. Their second design relies on people power for propulsion instead.

“It’s a passive device. The user pushes it. The handlebar is like the handlebar of a bicycle to a certain extent, except it’s spring-loaded. So the person can indicate their desired direction, such as forward, left or right. If that path is safe for the robot then the wheels will turn,” says MacNamara.

The onboard computer determines the directions in which a user can go, using video input to determine the prominent lines in the environment. The surrounding corridor is then classified based on those lines, and a verbal cue tells the user where they can go.

“We tried to do everything with voice because that is the most natural interface for an elderly person. Spatially, to remember where the Braille display might be is difficult for them, and a lot of these people go blind quite late in life and never learn to use Braille,” says MacNamara. “The user interface has to be as simple as possible. We only have one switch on it.”

Dubowsky says it’s that kind of user-specific information that makes clinical testing for both robots so invaluable. “The lesson to be learned from the field tests is: Don’t work with graduate students. They don’t give you right answers. You have to work with the elderly. You might think that the elderly do not accept new technology, but the ones we worked with were very comfortable with the system.”

MacNamara knows that cost is alsa an important factor for his product, and he hopes to keep it down to around the same price as a good electric wheelchair. “The idea would be to make the device affordable enough for insurance companies to say it’s worthwhile to install in their facility,” he says.

Another transportation method was explored by a team of researchers at Kanazawa University. A semi-autonomous system is used to navigate a wheelchair from one location to another. Although the concept is still in the early testing stages, the system will ultimately use ultrasonic beams placed throughout the facility. Acting as artificial landmarks, these beams let the wheelchair know its current position and guide it to its destination. The wheelchair also transmits an ID code to the central computer, so that the hospital or nursing home can keep track of the patients.

Kanagawa Institute of Technology and Kanagawa Technical High School in Yokohama, Japan, presented a novel idea for a “step-climbing wheelchair robot.” Designed to go over those little bumps at the edges of the road, this device might be better described as a curb climber than a stair climber, but its utility is undeniable. The wheelchair has specially designed rear wheels that slide forward to reduce the amount of power needed to lift front wheels. Research scientist Yoshihiko Takahashi calls this method “inverse pendulum control.” Once the wheel has moved into the forward position, the simple shifting of body weight is all it takes to help the chair over the curb.

MIT researcher Joe Spano explained that for some people, the problem begins before they even get into the chair: “We have certain patients that require assistance to get in and out of their wheelchairs and transfer to bed -- in some cases perhaps, manipulated onto the bed,” he said. “So we were interested in developing active services that could move the patient tangentially across the surface without the usual pulling and dragging.”

Spano’s transport system consists of a collection of rubber tips that look like upside-down ice cream cones with the pointy ends cut off. The extender tips move in tandem, up and down in a swinging motion that carries whatever’s lying on top of it like a leaf on a pond full of waves. “Supporting a human directly on a bed of extenders is suitable for shorter periods of time where you’ll be transporting a human, but for longer periods of time it is necessary to support the human on a hydrostatic air-filled mattress,” he says. A vacuum can be used to remove the air from the bed and expose the extenders for easy transport.

Using video clips, Spano demonstrated how to move a lot of things on those extenders. Wood, wood with a weight on top of it and Jello. Yes, Jello. Apparently, gelatin is a good substitute for an elastic substance like fatty tissue. The audience watched the video tape in awe as the extender sent the red block of goo down its length — first in a ziplock baggie, and then on its own.

He has also tested the device on a friend’s child, but it’s clear that the kid will soon outgrown the prototype, which is only a few feet long. There are still other hurdles to overcome, including modifying the extenders to turn patients around or move them across on angle. These are all things that existing conveyor belts can’t do.

With testing on all of these devices still incomplete, there remains much work before they will be available commercially. But, it seems clear that all of the traditional devices to care for the elderly, such as canes, walkers, wheelchairs and conveyor belts, will be greatly improved by the field of robotics.

Back to Clippings Index