A robot seal who’s helping dementia patients,
Meet Paro, an amazingly lifelike robot built to comfort people suffering from dementia. Although the automated baby seal has helped patients around the world, some worry it’s the first step on a slippery slope toward replacing human care
NANTO, JAPAN— In March 2002, a helicopter carried Takanori Shibata over the Gulf of St. Lawrence and deposited him on an ice floe, just off the coast of Quebec’s Magdalen Islands.
The Japanese engineer had spent the previous nine years developing “Paro,” a device designed to tackle one of the biggest public health issues of the 21st century: dementia. That day, bundled in a bright red snowsuit, more than10,500 kilometres from home, he was finally meeting the inspiration for his invention. “Mehhrrrr!” cried the baby harp seal. Very cute, the engineer thought, as his video recorder captured the moment.
Thirteen years later, that adorable seal cry has been programmed into 3,500 Paro devices deployed to more than 30 countries — everywhere from Japan, where Paro is manufactured, to the intersection of Queen and Ossington at Toronto’s Centre for Addiction and Mental Health. So what exactly is Paro? For starters, it is the spitting image of the Canadian baby seal Shibata met on that ice floe (and every bit as capable of provoking “awwws”).
But under its soft white fur, Paro is a robot — an artificially intelligent automaton that can flap its flippers, respond to its name and provide inexplicable comfort to people gripped by dementia.
The device is currently the subject of clinical studies in at least 10 countries — including Denmark, where local governments are buying Paro for nursing homes — and is an FDA-approved medical device in the United States. For many patients and caregivers coping with dementia, Paro has become an unexpected, and unusual, source of relief.
“He’s my best friend. I feel less afraid because he can listen to me.”
SUMIKO TOKUNO WOMAN, 80, SUFFERING FROM AND DEMENTIA
“He’s my best friend,” says Sumiko Tokuno, an 80-year-old woman with schizophrenia and dementia living in Nanto, 280 kilometres northwest of Tokyo. She beams as she strokes Paro. “I feel less afraid because he can listen to me.”
Tokuno is one of 47 million people worldwide living with dementia, an umbrella term that refers to a group of disorders affecting memory and cognition. Dementia comes in many forms but often features behavioural and psychological symptoms such as agitation, delusions and wandering.
These symptoms partly explain dementia’s enormous strain on caregivers. And as technology becomes more sophisticated — and healthcare workers become more overworked — some people are turning to robots for help.
On the market since 2005, Paro is one of the earliest robots to be used in dementia care. For people like Tokuno, the baby seal has proven remarkably effective.
When Tokuno arrived at Nanto’s Care Town, she was deeply paranoid, heard voices and was prone to nighttime wanderings. She barely spoke or ate, said Care Town director Sakii Takashi.
On a recent summer day, Tokuno is the happiest person in the room, chatting animatedly. She is calm and rarely wanders — all thanks to the seal robot, Takashi says.
“She felt Paro’s blinking eyes were agreeing with her and she got better,” he says. “And for her, now she has a responsibility: to look after Paro.”
Paro is astonishingly — and perhaps disturbingly — lifelike, thanks to its design and artificial intelligence. The seal looks at people when spoken to, cries in response to stimuli, and closes its eyes in apparent pleasure when scratched under the chin. Paro can even learn its name (in Australia, patients tend toward more familiar names like Bob or Mary).
The seal’s large, inky black eyes are oddly expressive. “I feel it’s looking into my soul,” a dementia patient once told an Australian researcher.
“I didn’t want to develop a robot that had emotional expressions,” says Shibata, who began work on Paro in 1993. “I wanted to develop a robot that (made people think) the robot had emotions.”
Shibata says Paro fulfils a similar role as therapy animals — but without the risk of bites or allergies, or the need to eat, sleep and poop.
Robots are tireless, with an endless supply of patience. Some dementia patients may have violent reactions toward therapy animals, but Paro remains indifferent when beaten or thrown across the room.
Yoshinobu Takeuchi, a social worker with Nanto City Hospital, says Paro is comforting precisely because it lacks thoughts and feelings. Agitation, he explains, is often linked with the feeling of being ashamed, judged or ignored.
“Paro just accepts the patient no matter what,” Takeuchi says. “Every question they have and every word they say, Paro reacts. And to be accepted is the most important thing in reducing anxiety and fear.”
Takeuchi says the medical establishment has been largely dismissive of Paro. “Almost everyone I talk to about Paro says, ‘That’s ridiculous,’ ” he says. Some ethicists also criticize robots as being deceptive, infantiliz- ing or the first step on a slippery slope toward replacing human care.
It doesn’t help that only a handful of studies have evaluated Paro and most are lacking in scientific rigour. But in Australia, Griffith University professor Wendy Moyle is hoping to provide some empirical answers.
Moyle was once a Paro skeptic — “I thought it was gimmicky” — but became a convert after conducting a small pilot study. “I was extremely surprised,” says Moyle, who observed that Paro made some people happier and less agitated or prone to wandering.
She is currently conducting the largest-ever clinical trial of Paro and is now analyzing the mountain of data collected by her team. Whatever the results, Moyle emphasizes that Paro isn’t a one-size-fits-all solution — but she does think this odd little robot can have a profound effect for certain people.
Moyle recalls one man who spent his days sitting in silence. When presented with Paro, his eyes widened. He picked the seal up and held it to his shoulder like a baby. Then when it was time for Paro to leave, he clung to its furry flippers and said something that brought everyone to tears.
“He said ‘goodbye,’ ” Moyle recalls with amazement. No one had heard him speak in two years. Helping people ‘recover a smile’ Inside a Korean conference centre, Takanori Shibata is setting up. He taps away on his laptop. He queues up a PowerPoint slide. Then he reaches into his duffel bag and pulls out a fluffy baby seal, which he strokes with the pride of a new dad.
At 48, Shibata has spent nearly half his life with “Paro” (a crude acronym for “personal robot”). Since 1993, the baby seal has taken over his life — occupying not only Shibata’s office (which is crowded with more than 40 of the robots) but also much of his time, taking him to speaking engagements around the world.
Shibata was born in Nanto, a city of 54,000 surrounded by low mountains and rice paddies. Here, people often wind up working in agriculture or the metals industry. But after he graduated from high school, Shibata turned to a more unconventional career.
“I thought that robotics would be a good future,” he says. “I wanted to have something special for me.”
After earning degrees in mechanical and electrical engineering, Shibata was hired at the National Institute of Advanced Industrial Science and Technology, where he still works as chief senior research scientist.
His interest was in creating automatons that can adapt and learn, as humans, animals and insects do. Shibata especially wanted to develop a “personal robot” that people could use in their daily lives.
He considered a robot to perform household chores, such as vacuuming and washing dishes, but soon re- alized each task would likely require a specialized machine.
Then he wondered: what if the robot fulfilled an emotional need, rather than a physical one? This got him thinking about pets. “People love them and appreciate their existence. But I wondered, ‘What is their role for human beings?’” said Shibata, who is not a pet owner. “They are mostly useless in terms of work — but animals can enrich our lives psychologically.”
Shibata decided to develop an animal robot to be sold in two versions: one for entertainment and one for therapy. He went for the obvious candidates: cats and dogs.
But after testing early prototypes, he had a revelation — these fourlegged creatures are just too political.
Dog lovers sniffed at the cat robots and cat people turned up their noses at the dog robots. And everyone was picky. It wasn’t enough to give a dog person a dog robot; it had to be their favourite breed.
Enter the baby harp seal. Soft and adorable, their bodies are also “eggshaped” and perfect for hugging.
Best of all, most people have never encountered a harp seal.
“Dr. Shibata has been very good at developing something that — certainly in most societies — people don’t have any negative reactions to, because they don’t know what it is,” says Moyle, the Griffith University professor. “When we first give Paro to people, everyone’s like, ‘What is it?’ And that’s part of the discovery, part of the interesting appeal.”
Shibata has spent two decades working on Paro, now in its ninth version. He has studied animal therapy and visited nursing homes. He distributed questionnaires to 2,000 people across seven countries.
Shibata, who is married with two children, spends nearly half the year travelling with Paro — he has visited Korea, Paris, Toronto and Los Angeles in the past few months alone — and introducing it at conferences or meetings.
He believes Paro can help people “recover a smile.” Shibata is still affected by the memory of an American nursing home he visited in the mid-1990s.
The elderly people were staring vacantly at the television. Nobody was smiling, everyone was silent.
“They were like . . .” he says, searching for the word, “machines.” All medical interventions should be scientifically evaluated, not only to ensure they work but also to establish that they are doing more good than harm. This is true even of robotic baby seals. And while Paro is unlikely to cause physical harm, important questions remain: What effect does it have exactly? What are the potential downsides? And is the roughly $8,000 cost really best spent on an animatronic harp seal?
Wendy Moyle, a professor at Australia’s Griffith University, is hoping to come up with some answers. The aging and dementia researcher is the principal investigator of the largestever clinical trial of Paro, which received about $1 million from Australia’s National Health and Medical Research Council in 2013.
The trial has 410 participants from 35 Australian facilities, all randomized into three clusters: a control group, a Paro group and a third group given a non-robotic version of the seal. “What I’m interested in is: is it the fluffy toy — or is it the actual robotic interaction that’s helping?” says Moyle.
The two Paro groups are allowed to interact with the seal in 15-minute sessions, during which they are observed and recorded on video.
Moyle was careful to remove any variables that might skew the results. This means, for example, that volunteers administering the study are asked to follow the same script when explaining Paro to patients.
Study subjects must also interact with Paro without any interference. “If you had a nice person giving the Paro and they’re interacting — then is it the person or is it the Paro (creating the effect)?” Moyle says.
Moyle’s group is now sifting through a mountain of data: they’ve interviewed friends and family about behavioural changes; they’ve recorded sleep patterns and used facial recognition software to analyze mood changes; they’ve even used sensor monitors to count people’s steps, so they can analyze whether Paro reduces wandering.
Moyle is particularly keen to learn whether Paro can reduce or delay the need for psychotropic medications. She also wants to evaluate whether Paro is worth its steep price tag. “Is this a cost-effective intervention compared to a plush toy or compared to usual care?”
Moyle says some patients have reacted negatively to Paro, throwing it across the room or simply ignoring it. She has also faced criticisms from her peers for studying Paro at all. “People say, ‘Why are you using technologies with people with dementia?” she says. “I often get ethics academics saying, ‘This is dreadful. You shouldn’t be doing this work.’ ”
But Moyle thinks it would be irresponsible to ignore technologies such as Paro, especially when it comes to dementia — an incurable disease that devastates its victims and places enormous physical, financial and emotional strain on their caregivers.
“Internationally, there are not enough people looking after older people and people with dementia,” she says. “And if technology can assist . . . and it engages people and makes them happy, then what is wrong with it?” Jennifer Yang