For Peg Fong Research Paper

1205 Words3 Pages

170224
LANGARA
Feature Writing
BONNIE LEE LA MADELEINE Assignment: Feature Story #2, Jun Tani For Peg Fong Emerging In that moment just before a soap bubble pops, its membrane will quiver frenetically. As Qrio, one of the humanoid robots that littered Jun Tani’s lab, puzzled out the difference between grasping a cup and moving it from side to side, the tension in the room felt like a soap bubble in its death throws. Jun Tani, a neuro-robotics scientist who was teaching Qrio, looked like he was going to pop. The problem seems simple. For months, Tani had been training Qrio to perform two tasks with his arms: 1) pick up a the mug, and 2) move the move from side to side. However, Tani had stripped the child-size robot of its original software …show more content…

He believes that robots with the right type of programming are ideal research models for understanding the brain: better, he argues, all those transgenic mice that garner headlines in the newspapers. He is trying to understand how a blank slate, with no knowledge of movement might start to understand …show more content…

A seven-inch stick with a ball on the top extended out from a platform on the table. The mechanism is connected to its brain via a cable, and Tani watches what happens in that brain on a monitor in the next room during a single training session and any experimental run. (The complete set of data are analyzed later, once the "brain" is downloaded to the computer in his office.) The joystick is taught how to move by an external teacher. Once the lesson is over, the joystick demonstrates what it has learned by executing, to the best of its ability, the learned movement. This is not a programmed path that the robot takes, it is a path that the robot chooses based on its understanding of what it has learned. This is a fundamental difference between the robots at Disney and that joystick. The joystick never gets the movement it was taught exactly right. It gets close, but never exact. However, two other key observations emerged from his experiments with this robot that helped him see that he was heading down the right path to creating a robot with a human-like learning

Open Document