Blog: Automaton, Know Thyself: Robots Become Self-Aware

Automaton, Know Thyself: Robots Become Self-Aware
Scientific American (02/24/11) Charles Q. Choi

Cornell University's Hod Lipson is developing robots that can reflect on their own thoughts by equipping them with two controllers. One controller was rewarded for chasing dots of blue light while avoiding red dots, and the second controller modeled how the first behaved and how successful it was. This technique, known as metacognition, enabled the robot to adapt after about 10 physical experiments, as opposed to the thousands of experiments needed using traditional evolutionary robots. "This could lead to a way to identify dangerous situations, learning from them without having to physically go through them--that's something that's been missing in robotics," says University of Vermont's Josh Bongard. Lipson also is studying how robots can model what others are thinking by programming one robot to watch another randomly move toward a light. The observer developed the ability to predict the other's movements so well that it could lay a trap for it on the ground. "This research might also shed new light on the very difficult topic of our self-awareness from a new angle--how it works, why, and how it developed," Lipson says. One application for self-aware robots could be the maintenance of a bridge, with sensors constantly monitoring vibrations in the framework to develop a self-image of the bridge.

View Full Article