Automation Time-Saver: Teaching Robots to Teach Other Robots New Tricks


Most robots are programmed using one of two methods - learning from demonstration, in which they watch a task being done and then replicate it, or via motion-planning techniques like optimization, which require a programmer to specify a task's goals and constraints. Robots that learn from demonstration can't easily transfer one skill they've learned to another situation and remain accurate. The other method can be time-consuming since it requires hand-coding by programmers. MIT researchers introduce C-LEARN, which allows non-coders to teach robots a range of tasks simply by providing some information about how objects are typically manipulated and then showing the robot a single demo of the task. This enables users to teach robots skills that can be automatically transferred to other robots that have different ways of moving - a key time- and cost-saving measure for companies that want a range of robots to perform similar actions. The team tested the system on Optimus, a new two-armed robot designed for bomb disposal that they programmed to perform tasks such as opening doors, transporting objects, and extracting objects from containers. In simulations they showed that Optimus' learned skills could be seamlessly transferred to Atlas, CSAIL's 6-foot-tall, 400-pound humanoid robot.