Evolving a Behavioral Repertoire for a Walking Robot
This addresses the limitation of existing walking algorithms for robots in real-world missions by providing a more versatile behavioral repertoire, though it is incremental as it builds on evolutionary methods.
The paper tackles the problem of enabling legged robots to learn diverse walking directions beyond straight-line motion, resulting in a repertoire of controllers that allows a hexapod robot to reach every point in its reachable space with only a few dozen short experiments.
Numerous algorithms have been proposed to allow legged robots to learn to walk. However, the vast majority of these algorithms is devised to learn to walk in a straight line, which is not sufficient to accomplish any real-world mission. Here we introduce the Transferability-based Behavioral Repertoire Evolution algorithm (TBR-Evolution), a novel evolutionary algorithm that simultaneously discovers several hundreds of simple walking controllers, one for each possible direction. By taking advantage of solutions that are usually discarded by evolutionary processes, TBR-Evolution is substantially faster than independently evolving each controller. Our technique relies on two methods: (1) novelty search with local competition, which searches for both high-performing and diverse solutions, and (2) the transferability approach, which com-bines simulations and real tests to evolve controllers for a physical robot. We evaluate this new technique on a hexapod robot. Results show that with only a few dozen short experiments performed on the robot, the algorithm learns a repertoire of con-trollers that allows the robot to reach every point in its reachable space. Overall, TBR-Evolution opens a new kind of learning algorithm that simultaneously optimizes all the achievable behaviors of a robot.