CNET también está disponible en español.

Ir a español

Don't show this again


You're not playing Angry Birds anymore, but robots are

A new study from Georgia Tech shows that kids like to teach things to robots more than they do to adults. And that's OK.

Researchers ask kids to team up with robots and avians to vanquish the dastardly pigs in Angry Birds. Video screenshot by Michael Franco/CNET

Many parents might fret about the hours their kids spend swiping their fingers across a tablet's screen. But for children with developmental or physical challenges, such an activity could serve as valuable therapy.

One way to get kids to engage in the repetitive motions they may need to increase their motor coordination skills is to have them teach a game to someone else.

But when researchers at Georgia Tech asked children to teach Angry Birds to adults, they stayed interested in the activity for only 9 minutes. When they paired up a robot with an Android tablet, though, and asked the kids to explain the game to their electronic buddy, the kids engaged three times as long and made eye contact with the robot 40 percent of the time, as opposed to just 7 percent with the humans.

"Imagine that a child's rehab requires a hundred arm movements to improve precise hand-coordination movements," Georgia Tech's Ayanna Howard said in a statement. "He or she must touch and swipe the tablet repeatedly, something that can be boring and monotonous after awhile. But if a robotic friend needs help with the game, the child is more likely to take the time to teach it, even if it requires repeating the same instructions over and over again. The person's desire to help their 'friend' can turn a 5-minute, bland exercise into a 30-minute session they enjoy." Howard, a Motorola Foundation professor in the School of Electrical and Computer Engineering, is leading the project.

In addition to helping kids overcome challenges, the study showed that the training activities can help robots too.

During the sessions, the robots sit on a desk next to the child and watch the young ones' maneuvers, recording them in their memories. The robots are programmed to observe starting and ending finger positions and to monitor the scores for signs of success. The machines then take their own turn at bird flinging and are programmed to do a little happy song (which sounds an awful lot like R2-D2) and dance if they win. Conversely, they slump a bit and shake their heads if their bird is a dud. In effect, the robots learn how to interact with both the game and their human teachers.

Such learning ability, the researchers feel, is vital to speeding up the integration of robots into everyday lives.

"One way to get robots more quickly into society is to design them to be flexible for end users," said Hae Won Park, Howard's postdoctoral fellow working on the project. "If a robot is only trained to perform a specific set of tasks and not able to learn and adapt to its owner or surroundings, its usefulness can become extremely limited."

Park and Howard feel that such training flexibility could allow them to, for example, quickly and easily program a robot to work with a child on a specific set of tasks and then let the child take the machine home where it would work as programmed and also adapt to his or her behaviors.

Next up for the researchers is to have children with Autism Spectrum Disorder (ASD) teach the fun-loving robots new games including Candy Crush and ZyroSky while they monitor the therapeutic effects of the activity.