Robots have so far generally been bad at adapting to new situations, such as recognizing new objects or coming up with their own ideas about how to carry out tasks in changing environments. One solution is to equip them with endlessly detailed instructions to minimize the amount of unfamiliar things they experience. Another, more elegant option, is to teach robots to think and adapt for themselves.
What’s the Big Idea?
Cornell’s Personal Robotics Laboratory is teaching a robot to generalize groups of objects, which is one of the most basic aspects of reliable adaptability. For example, instead of teaching robots “this is a cup, and this is a slightly different cup” and so on, they have found a way to teach them to recognize features common to all cups. Thus, when they see something cup-like, they can say to themselves, “hey, that’s a small container with a handle, I bet it’s a cup!”