From building cars to assisting in high-precision surgeries, robots today are taking on more complex roles than ever before — and getting smarter and more sophisticated every day.
With recent advances in programming algorithms and artificial intelligence, the possibility of robots moving from the factory floor into our homes — and even looking after our children — is a fast-approaching reality. Think Rosie, the space-age robot maid and nanny to “The Jetsons.”
Some of these robots are already being developed. Consider Pepper, a 4-foot-tall interactive robot able to recognize human emotional states by analyzing voice tone, facial expressions, and other non-verbal cues. Or Jibo, an 11-inch-tall personal assistant, storyteller, and communication device being marketed as the first "family robot."
Despite these clever androids, we're still far from having Rosie-like robots we can trust as nannies. Today's autonomous bots still lack the manual dexterity to, say, pour a child a cup of milk, or the emotional instincts to soothe a crying toddler. "Having a robot that's capable of those types of things won't come for the next 20 or 30 years," says Henny Admoni, a roboticist at Carnegie Mellon University.
That's not to say today's robots serve merely as entertainment devices. Admoni and other experts see these machines playing a significant role in children's emotional, social, and cognitive development in the near future.
There are several concerns around giving robots these responsibilities, including worries about privacy and about what unintended effects robo-tutors will have on kids' development. Still, research suggests that social robots will be able to enhance learning by helping preschoolers recognize and resolve conflicts, teaching children with special needs the social skills to successfully interact with others, and assisting school children with their math, grammar, and language skills.
"We really are at a transition point in human-robot interactions," Admoni says.
Overall, research shows that children can benefit from interacting with robots, but it's important to recognize that these benefits are less pronounced than those a child would get from interacting with a person, says Solace Shen, a Cornell University psychologist who studies robot-human interactions, particularly in the development of children. "The goal is not to have the robot replace interactions with humans,” she says. “But more to supplement them."
Assisting preschool teachers may be one of the first tasks child-friendly robots take on.
In an on-going research project, Shen and her colleagues are considering the role robots can play in groups of children age 3 to 6. In preschool classrooms, teachers are often unable to focus on every conflict among students, Shen says, adding that the number of conflicts in this age group is quite high — about six to nine per hour — and often results from issues with sharing.
Using a "Wizard of Oz" approach, in which the researchers secretly control a robot's actions, Shen is finding they can ameliorate these situations by helping children first understand the conflict. "Once they can acknowledge there is a problem, they can use the conflict-resolution skills they're already being taught to resolve things and come up with ways that everyone gets what they want," she says.
(Previous Cornell research found that robots can similarly help adults work better in teams simply by pointing out when someone is using harsh, destructive language).
Placed in preschool classes, social robots like Pepper could use their powerful emotion-recognition engines to spot minor squabbles, which would allow teachers to focus on the larger meltdowns that occur. The robots may also be programmed with established negotiation strategies to better resolve conflicts and further reinforce skills children are developing.
Robots can also help improve the emotional and social development of children with special needs, such as those with autism or Down syndrome because these machines have several characteristics that make them attractive to these children.
"We really are at a transition point in human-robot interactions."
For one thing, a vast body of research shows some kids with autism respond well to technology in general, including computers, phones, tablets, and robotic toys. Studies also suggest robots are appealing to special needs children because they're less complex and more predictable than people, less intimidating, perpetually patient and consistent in the tone of voice and mood, and highly customizable and adaptable to children's specific needs.
While some autistic children have difficulty maintaining eye contact, the same isn't always true when they interact with robots, adds Jason Borenstein, a bioethicist at Georgia Tech whose research focuses on robotic caregiving to children and the elderly. "For whatever reason, they tend to bond more quickly with a robotic entity," he says.
For these reasons, researchers have used robots to engage with special needs children and elicit numerous behaviors, including initiating interactions, imitating behaviors, learning to take turns, recognizing emotions, and focusing their attention.
One oft-used robot in this research is Kaspar, a child-sized droid that’s comfortable for autistic children to interact with because of its simplified speech, gestures, and facial and body expressions. In some studies, researchers allowed isolated autistic children — those who don't respond to or interact well with humans — to play with Kaspar while a teacher or experimenter was nearby. In one case, a child touched his teacher's face and eyes after excitedly exploring Kaspar's face and later invited the teacher to join in a game with Kaspar.
"The benefit of robots is that we can bootstrap child-technology interactions in order to lead to better child-adult and child-child interactions," Admoni says. In essence, Kaspar and other robots act as mediators, allowing special needs children to apply what they learn from robots in their interactions with other people.
Such robots are slowly making their way out of the research lab and into the world, including Kaspar, which researchers are looking to redesign and commercialize. Another robot, Leka, which was financed through a crowdfunding campaign, is designed specifically for kids with autism. The small, ball-shaped bot is equipped with facial expressions, lights, colors, and games, including picture matching and hide-and-seek. Leka was designed, in part, to be an interaction bridge between children and their surroundings (including caregivers and other children) and has reportedly been shown to facilitate play between children with cognitive disabilities.
Aside from engaging with children on the social and emotional level, these robots will soon enrich children's cognitive skills, particularly at home. "Robots will potentially help children with things like math problems and learning to read," Borenstein says.
Robots will also be helpful in children learning languages because they will allow a degree of language immersion at home that children don't normally receive in the classroom, Admoni says. What's more, research shows robots, such as the furry, literacy-minded Tega, can help preschool age children improve primary language skills and build greater vocabularies through storytelling activities.
In part, robots are great tutors for language and other primary school subjects because they're the epitome of patience. "Even the most well-intentioned and kind teacher, mother, or father has a limit," Shen says. "Once you repeat something for the 10-millionth time, you're not going to want to do it again. But robots don't have that problem and children love the repetition."
Somewhat paradoxical, robots are sometimes more effective tutors when they're playing dumb. Scientists in Japan found that children learned English vocabulary words better when robots made mistakes and the children had to correct their mechanized study partners, likely because doing so boosts self-confidence and reinforces existing knowledge.
Robots can potentially take their tutoring lessons to places human teachers may not be able to, such as isolation units in hospitals. In these cases, Shen says, robots can help ill children keep up with their studies, and provide emotional and social support when parents aren't able to be there.
Are We Ready?
Despite the wealth of potential benefits robot tutors present, there's much to consider.
From an ethics standpoint, one potential concern with social robots is privacy. If robot tutors have the ability to record their sessions, who should those recordings be shared with and should children know they're being recorded? Additionally, robots can mimic emotions but don't actually feel them — is it ethical to let children believe robots are happy to spend time with them?
Though there's little evidence that interactions with robots will stunt children's emotional and social growth, some experts are concerned that children may develop a kind of master-servant relationship with robots that then translates into their interactions with people. That is, if robots are programmed to follow orders and are unable to experience pain and explain that feeling to children that may bully them, will this affect what children believe to be socially acceptable behaviors?
There's also a potential issue with our implicit biases and stereotypes making their way into the design of the robots (reflected in the way it answers questions). This may be amplified if the androids have access to the internet.
"If you ask a robot to describe a physician and it says that a physician is a male that treats patients, then that's an issue," Borenstein says. "It means our practices and biases are being reflected in artificial form."
With robo-tutors likely hitting the market within the next several years, these concerns won't be put to rest before the robots are in children's hands. Yet, while these issues are important to consider, various social and communicative technologies — including robots — are generally improving people's lives rather than harming them, Admoni says.
"Most people designing robots are really looking to fill a void that already exists," Admoni says. "We're building robots that are not replacing people but are helping in new ways to improve children's learning. It's a tremendous time for human-robot interactions."