Improvement of learning efficiency by exploiting multiarticular muscles -a case study with a 2D serpentine robot-

This study is intended to deal with the interdependency between control and body systems, and to discuss the "brain-body interaction as it should be" particularly from the viewpoint of learning, by borrowing the idea from the "protein folding problem". As a practical example, we demonstrate decentralized control of a 2D serpentine robot consisting of several identical body segments. The preliminary results derived indicate that the convergence of decentralized learning of locomotion control can be significantly improved even with an extremely simple learning algorithm, i.e., a gradient method, by introducing biarticular muscles compared to the one only with monoarticular muscles. This strongly suggests the fact that a certain amount of computation should be offloaded from brain to its body, which allows robots to emerge various interesting functionalities.