Could Robots Help Kids Learn Second Languages Faster?

European researchers are testing a robot language tutor that will give refugee children one-on-one attention.

Ideally, every child learning a new language would have their own teacher. That sort of attention helps students learn more quickly. But, of course, most schools can’t afford one-on-one learning, at least not for the whole classroom.


Could robots help fill in the gaps where human teachers can’t be?

That’s the question a big new research project in Europe is trying to answer. Funded by the European Union, L2TOR (pronounced “el tutor”) aims to develop robots that speak directly to students and respond to their emotional cues. If a child is bored, the robot will adapt its approach. If it senses contentment, it will keep going.

“This is aimed at children aged four to six,” says James Kennedy, a research scientist based at Plymouth University, in the U.K. “The robot might be in their homes, or at their preschool or kindergarten. Humans will be around, but the robots will be the ones delivering the material in both the child’s native language and the language they are trying to learn.”

Five universities and two companies are collaborating on L2TOR, which uses the NAO humanoid from a French company. Researchers hope to help the kids of recent immigrants, like Turkish children arriving in Germany or the Netherlands for the first time.

“We’re looking for the robots to add something to the human experience,” Kennedy says. “Typically, you have a high ratio of children to adults, so adults can’t provide as much one-to-one time as they like. Giving some of that time over to a robot adds to what [the students] are getting. We’re not replacing humans in any way.”

Kennedy’s team has already used robots in local schools, teaching the genders of French words. Using an accompanying touchscreen, kids pressed a word in English, then the robot said the French word with either a “le” or a “la.” “The children learned very quickly, picking up both the gender and a lot of vocabulary. It’s tremendously exciting for them,” Kennedy says.


One question the research will have to answer, though, is whether kids learn because of the excitement of using robots, or because using robots is a fundamentally good way to teach students. Only long-term observation will tell.

The NAO robots are preprogramed with text-to-speech ability, but the researchers may have to record certain words so they sound authentic. “We need to look into what accent the kids pick up because we wouldn’t want them to go around with a very robotic accents,” Kennedy says.

All Images: via L2TOR


About the author

Ben Schiller is a New York staff writer for Fast Company. Previously, he edited a European management magazine and was a reporter in San Francisco, Prague, and Brussels.