Real-time Control of Bipedal Movement based on Basal ganglia and Brainstem systems

During the bipedal locomotion in the real world, the body will receive unpredictable forces depending on various factors. Because the biological system has an ability to change various dynamical properties of its own body, an appropriate control of those properties is important for the system to maintain the locomotion in the real world. Muscle-tone is one of important parameters to determine the properties of the body, such as the stiffness and the viscosity. To adapt unpredictable changes of the environment, controlling the muscle-tone seems to be one of useful ways. We show results of model simulations, indicating that flexible and robust bipedal locomotion can emerge from appropriate control of the muscle-tone depending on the ground reaction force. Thus, by determining the properties of body dynamics with adapting unpredictable environments using ground reaction forces, model could be adapted to various loads without needing an explicit posture control