It seems if you fill a classroom with robots that make mistakes, the kids get smarter. According to this New Scientist Report, a Nao Robot was operated by humans in the next room during an English class in a Japanese school. Yes, that is Telepresence rather than true robotics, but the kids didn’t know that. They played a learning game where the English name for a shape was given, and the robot and kids would draw that shape. It appears the kids learned faster when the robot made mistakes, and the children would have to teach it to draw the correct shape to go with that word. Which is just scientific backing for the old adage The best way to learn is to teach. Not only that, but the kids then wanted to continue learning with the Robot, and would carry on studying longer and learning better as they did so. The results will be presented at Ro-Man this year, the 21st IEEE International Symposium on Robot and Human Interactive Communication, which will run from tomorrow, September 9th through the 13th in Paris, France. The event is all about real world results of humans and robots working and communicating together, and every year contributes tremendously to the further development of robotics on both the software and hardware fronts. Take a look at their scheduled presentations to get an idea of the scope of this event. If you were thinking of building your own robot, this is a great place to absorb some real understanding of what is possible today and coming for tomorrow.