Body intelligence: how passive mechanics shape the motor control 
of human locomotion

The talk explores the mechanical principles of legged locomotion, and 
their consequences for the control of the human leg and of artificial 
legs in prosthetics and rehabilitation. First, I will use simple 
spring-mass models to unite the two fundamental gaits of walking and 
running in one theory, and to demonstrate the importance of mechanical 
self-stability in legged locomotion. Second, I will shift from simple 
mechanical to neuromechanical models that embody the spring-leg behavior 
in their muscle-reflex control. I will present a hopping model, which 
shows that a usually fatal, positive force feedback is key to respond 
spring-like in muscle-driven legs. Third, I will extend this motor 
control idea to a complex biped model of human locomotion, and show that 
neither central rhythm generators nor desired joint trajectories are 
required to walk and almost run, and to manage random ground and stairs. 
Finally, I will present current efforts to apply this powerfully 
flexible and self-adaptive reflex control to active ankle prostheses and 
to the neuro-rehabilitation of stroke patients.