Paro is a robotic baby seal. It may look like a toy, but it’s quickly attracting the serious attention of rehabilitation researchers. In Japan, more than 1000 units have been sold to care providers in nursing homes and hospitals, as well as to consumers who want a robotic companion. Short-term experiments in Japan and the United States show that Paro can have positive effects on the mental health of some elderly people. Now long-term studies are under way in Europe. The results could lead to specialized versions of Paro to help specific groups of people, such as elderly individuals suffering from dementia or children with autism.
It’s easy to see Paro’s attraction. It squeals, blinks, moves its head and paddles, and squeals some more. Oh, and it loves to be petted. But Paro—invented by Takanori Shibata, a researcher at Japan’s National Institute of Advanced Industrial Science and Technology (AIST)—is also smart. Artificial intelligence software changes the robot’s behavior based on a host of sensors that monitor sound, light, temperature, and touch. Paro learns to respond to words its owner users frequently. And if it’s not getting much petting time, it will cry.
Intrigued by these capabilities, researchers at the Danish Technological Institute’s Centre for Robot Technology began the first long-term study of Paro’s potential in elderly care. The researchers distributed 30 units to residents of nursing homes with various levels of senile dementia. The study will continue for several years, says Lone Gaedt, the project’s leader, but it’s already evident that Paro not only makes patients feel better but can also help them communicate better with others, including caregivers.
”You see people who had lost language pronouncing words or talking to Paro as if it was a pet they had in the past,” she says. ”You even see very debilitated people who can’t take care of themselves but want to take care of Paro.” But, Gaedt adds, the little robot is just one of many tools that should be used and that some people don’t have the mental resources the robot requires—or they just lose interest.
In the United States, where Paro will be commercially available sometime this year—at a cost of about US $6000 each—small trials are in progress at institutions like Vinson Hall, a retirement community in McLean, Va. Marcia Twomey, director of development, says that residents become very animated by petting and talking to Paro, much as they would with a real pet, but with the advantage that the robot lets users engage with it as long as they like and won’t walk away.
”Paro has made many friends here,” she says. ”It’s a very magnificent piece of robotics.”
AIST’s Shibata has performed his own tests with Paro users. He’s watched hundreds of hours of recordings of people interacting with the robot, and he’s also measured hormones linked to stress and analyzed brain activity. He says Paro produces a positive psychological and physiological effect on people. ”But so far Paro is more like a general-purpose therapeutic tool,” he says. ”To improve the therapeutic effect, we need to create specialized versions.”
The Denmark study and tests at places like Vinson Hall are helping Shibata understand how different types of users respond to the robot. His goal now is to fine-tune Paro’s behavior—and create new behavior—to enhance these therapeutic effects. For example, he’d like to make a version of Paro that would stimulate more verbal communication in demented patients who are losing their language skills. Another version would try to elicit more interaction between autistic people and their caregivers.
It will take several years to develop these new versions, Shibata says. But after working on Paro for nearly two decades, the researcher says he’s not tired of his, well, pet project. He enjoys Paro’s company, but he keeps the robots at the lab and has none at home. ”I have one in my car,” he says, ”just in case I need to show it to someone.”