Learning to control complex tensegrity robots

Abstract

Tensegrity robots are based on the idea of tensegrity struc-tures that provides many advantages critical to robotics such as being lightweight and impact tolerant. Unfortunately tensegrity robots are hard to control due to overall com-plexity. We use multiagent learning to learn controls of a ball-shaped tensegrity with 6 rods and 24 cables. Our simu-lation results show that multiagent learning can be used to learn an efficient rolling behavior and test its robustness to actuation noise.