Pitt, CMU researchers are seeking innovative ways to help people remain independent
November 14, 2007 5:00 AM
Research assistant Joseph Olson runs through a computer simulation of the task movements a wheelchair with a robotic arm must perform to open a refrigerator, select a drink container and open it.
University of Pittsburgh professor Rory Cooper explains how the staff at his wheelchair laboratory at the Highland Drive VA Medical Center will attempt to employ a robotic arm attachment that will allow a wheelchair user
to recline and lift to change a light bulb. Dr. Cooper's lab is working on applying computer and robotic technologies to wheelchairs to make them more useful for disabled people and older adults.
By Gary Rotstein Pittsburgh Post-Gazette
For now, a robotic wheelchair viewed as a future all-purpose assistant for disabled and elderly individuals is in pieces in different rooms of the University of Pittsburgh's Human Engineering Research Laboratories: a motorized chair here, aluminum arms with end-claws over there, computer sensors elsewhere.
By March, its developers hope to put the components together to have it open a door for a wheelchair user.
In 10 years, they want it to prepare an omelet for that person.
That kind of dramatic leap, from navigating passageways to frying eggs, represents just one aspect of a wide range of advanced technology research taking place jointly at Pitt and Carnegie Mellon University to help people stay independent.
The universities received a $15 million, five-year grant from the National Science Foundation a year ago to develop their new Quality of Life Technology Center, with similar funding expected to follow for another five years.
The idea is that after a decade, different teams of robotic scientists, mechanical engineers, clinicians and others will have performed "transformative" research. It is to create new possibilities for people trapped today in weak bodies or in residences and vehicles intended only for able-bodied individuals.
If the researchers succeed, and if economic realities permit whatever they develop to be mass-marketed, tomorrow's swelling ranks of longer-living, frailer individuals could remain productive longer and postpone nursing home stays. Some type of artificial assistance, be it a science fiction-style personal robot or far more subtle sensors, could physically help them dress, dine and drive or at least coach them in such tasks.
Both consumer representatives and officials from businesses interested in manufacturing the type of products that might be designed attended the first Quality of Life Technology Summit in Oakland last week to share ideas about the research under way.
Jim Osborn, executive director of the Quality of Life Technology Center, recently told a gathering of long-term care providers that if such advances could delay all nursing home admissions by a month, societal savings could be $1 billion monthly. Artificial assistance is necessary because a shortage of both paid personal attendants and available family caregivers is anticipated as the elderly population doubles over the next 30 years.
"We have lofty objectives," Mr. Osborn said of the center's work. "We have to do research that will turn the world on its head."
The robotic wheelchair work of a team headed by Rory Cooper, chairman of Pitt's School of Rehabilitation Sciences and Technology, is among the easiest of the center's 22 funded projects to visualize now, but the overall scope is far broader. About 50 faculty, staff and graduate students, evenly split between the two universities, are trying to accomplish a variety of high-tech goals. Among them:
• Developing systems that observe people's ability to perform basic functions in their homes, recognize when they decline in those abilities, and then call attention to it so that person -- or others monitoring by computer -- can arrange needed help.
• Inventing robotic devices that can go beyond recognizing deteriorating abilities to actually help people with tasks such as getting out of bed, getting dressed, preparing meals and following daily hygiene.
• Creating in-car programs to keep people mobile by observing their driving ability, detecting potentially dangerous habits and guiding them on how they can stay behind the wheel safely.
Those efforts follow on the heels of innovations such as remote medical screenings via connections to computers outside a person's home; vacuum cleaners that operate themselves; audible GPS navigation guides in cars; and robotic vehicles like the one Carnegie Mellon used to win the Defense Department Urban Challenge race in California on Nov. 3.
The difference between those examples and the objective of the Quality of Life Technology Center is the level to which local researchers want some version of a computer or robot to study humans, interpret their behavior, recognize their patterns and anticipate what they should do next. A variety of sensors, cameras, accelerometers, gyroscopes and more is being put into use in laboratories across Oakland to help figure such things out.
Creating a computerized system simply to record observations about human behavior is itself a big step, said Carnegie Mellon Robotics Institute professor Martial Hebert, and going beyond that is even tougher.
"Once you try to understand the real world and real actions of people and things like that, you get to a level of complexity very difficult to deal with," he noted. "Even something as simple as recognizing that somebody is standing up -- think of the billions of different ways that this can appear, so how can you capture that in a system and recognize that?
"That's what makes perception difficult [for a computer], the incredible variability of the environment," said Dr. Hebert, who is leading the work on perception and awareness issues with the center's director, Carnegie Mellon computer science and robotics professor Takeo Kanade.
In order for Dr. Cooper's wheelchair team to get its long-armed robot to the door it will open, it needs the Hebert-Kanade team to provide the sensory ability for a computer to recognize what a door looks like, perceive how its handle works, and then approach and use it. The ultimate idea is to remove the need for the person in the chair to operate the controls -- he or she might instead use a laser pointer to direct the wheelchair on where to go.
The center has partnerships with long-term care providers and disability service agencies in order to get feedback from their clients on the obstacles they face and concerns about the devices they're developing.
Breakthroughs via the technology can't come fast enough, said Dan Rossi, associate executive director of United Cerebral Palsy of Pittsburgh.
"There are so many physically disabled but cognitively aware, very bright individuals," Mr. Rossi said. "They're just locked in these bodies that don't work, so these technologies that can free them up to express their thoughts and feelings, to go where they want go and do what they want to do, those are just amazing things."
Cleo Cabuz, vice president of engineering for Honeywell Life Safety, which is already deeply involved in the tele-health business, said the company is intrigued by the idea of in-home devices that interpret a person's ability to perform activities of daily living.
"Is the person getting out of bed and doing the normal routine, having an adequate level of activity, or is that increasing or decreasing? Something like that reflecting on the general well-being of a person is very important to us, and it's important that it can be done in a nonintrusive and economic way," she said.
The cost of the potential products to consumers, though not the first thing the researchers are concerned about, is significant enough that the center has business advisers. It also has social scientists who assess the willingness of people to interact with robotic devices of various kinds and take instructions from them. The researchers spend time discussing how tactful verbal advice must be, such as when prompting an elderly motorist about his driving.
"We're seeking a trusted adviser type of approach, which is much more acceptable than something that just gives you a hard time," explained Aaron Steinfeld, a Robotics Institute systems scientist spearheading the safe-driving project.
Progress on the center's various projects is to be demonstrated to National Science Foundation officials in March. They will make a site visit to Pittsburgh, as one of 22 engineering research centers around the county receiving similar funding levels for entirely different "transformative" research, such as anticipating earthquakes and tornadoes. After three years, sufficient progress must be shown for the foundation to allocate the second five-year funding stream on top of the first $15 million.
If the project works as envisioned, the center's achievements will include "virtual coach" technology, in which a person with fading mental skills could wear a watch or carry a cell phone that registers all kinds of information. This mobile assistant might advise that person to take medications, watch a favorite TV show or shift in the wheelchair to avoid pressure sores.
The toughest part, researchers agree, is designing a robot to put its "hands" on humans, helping brush their hair or teeth, button their clothes, lift them from bed or make other advances envisioned to help the most disabled people.
"We're interested in a unified system that solves all these problems, ideally, for someone who's paralyzed and living alone," said Carnegie Mellon robotics professor Chris Atkeson. "The question is how can we make manipulating devices much more like a human hand or arm, very powerful but still soft. ... One of the things you have to be sure of is you're not going to hurt the human. If you break a dish [rather than a person], that's not the end of the world."
Mr. Osborn said it only makes sense for the elder-swelled local region to lead national efforts in attempting such technology. If the center does its job well, it will outlive the 10-year science foundation grant money and become self-sustaining by other means, with business spinoffs to manufacture or market the breakthroughs.
"We have a wonderful opportunity in a very genuine way to marry what the university research community is doing with the needs of the community at large," he said.
"It has pretty advanced underlying technology, but at the end of the day, everybody can appreciate what it's for. ... If there's going to be a cluster of companies that end up doing this, there's every reason in the world it should be here."