Brain-Inspired Motion Programming for Hobby-Use Humanoid Robots

Dynamic motion of hobby-use humanoid robots is defined by a sequence of static poses. Each pose is defined by specifying joint angles numerically for each joint. This is a difficult task, because typical hobby-use humanoid robots have more than 20 joints. We need a way to program robot motions easier. Our brain employs a different strategy for motion programming. Instead of specifying each joint angle or each muscle tone, the primary motor cortex has a set of primitive motions called motor primitives, and higher cortical areas and the basal ganglia generates a sequence of combinations of such motor primitives. Thus, our brain uses more abstract form to compose dynamic motions. In this article, we employ self-organizing maps (SOMs), which is an unsupervised learning algorithm used in machine learning and computational neuroscience communities, to extract motor primitives from a set of motion data of a hobbyuse humanoid robot. Then, we reprogram the original dynamic motions by the sequence of combinations of extracted primitives. We are able to reproduce the same motions by the proposed method. These results suggest that our brain-inspired method provides an easier means to program hobby-use humanoid robots than existing software bundled with the robots.