Achieving natural behavior in a robot using neurally inspired hierarchical control

Abstract

Terrestrial locomotion presents tremendous computational challenges on account of the enormous degrees of freedom in legged animals, and the complex and unpredictable properties of the natural environment and the effectors. Yet the nervous system can achieve locomotion with ease. Here we introduce a quadrupedal robot capable of goal-directed posture control and locomotion over rough terrain. The underlying control architecture is a hierarchical network of simple negative feedback control systems inspired by the organization of the vertebrate nervous system. Without using an internal model or feedforward planning, and without any training, our robot shows robust posture control and locomotor behavior in novel environments containing unpredictable disturbances.