Exploring the neuromuscular control of locomotion using a muscle-powered robotic limb

For decades, the neural control of vertebrate limbs has been heavily explored by neuroscientists and engineers. Accordingly there exist many mathematical motor control models that account for neural properties, but neglect (or oversimplify) the interplay between neural control and muscle tissue mechanics. Similarly, ‘bio-robotic’ models based on animals can measure external forces and motions, but offer no insight into internal muscular forces [1]. Consequently, it remains largely a mystery how the central nervous system controls the motions of multiple joints in a limb to achieve efficient and stable locomotion. Furthermore, there exists a dilemma in evolutionary biomechanics: we do not understand whether an animal’s impressive speed, stability or efficiency is due to their limb architecture or to the intrinsic properties of the muscle-tendon tissues (or some unknown combination). Using running birds as a model [2, 3], the aims of this project are a) to elucidate how neuromuscular control is achieved among the leg joints, and b) to decouple tissue vs. structural influences on locomotor performance. Using recently developed ‘musculo-robotics’ techniques [4, 5], these aims will be achieved in three broad objectives: 1) Build a muscle-actuated quail robot to mimic both biologically-realistic internal and external forces. 2) Experimentally manipulate the muscle tissue properties of this robot to address how muscle speed, strength and elasticity may enable locomotor stability capacity beyond what conventional theory or robotics predict. 3) Experimentally manipulate the musculo-skeletal anatomy of the limb (architecture) to address the relative importance of tissue vs. structural properties in determining how effectively a limb moves.