Joseph Sammartino Graduates with MS

Congratulations to Joe Sammartino for successfully defending his MS project about modeling the passive properties of the robot and rat hindlimbs.

Abstract

Animals perform locomotion over different timescales and different limb lengths. These limb lengths and
time scales play an important role in determining if the forces during motion are dominated by visco-elastic
properties, inertial properties, or damping properties. To explore how these forces affect neural control, a
framework for adding visco-elastic components to an existing quadruped robot (Muscle Mutt) is presented.
A mathematical model for three segment pendulum motion was derived using Lagrange-Euler equations of
motion. The model was then used to find joint spring and dampers for rat data and Muscle Mutt data. The
spring and damping constants were set by optimizing parameters to match free-hanging experiments of live
animals. Future work is needed to complete this framework, as problems with Muscle Mutt hardware need
addressing. Once these problems are fixed, legs can be designed by scaling the damping ratio and natural
frequency of the hind leg of Muscle Mutt to match that of a rat by adding a spring and damper at each joint.
This new leg will be used in future research to test the hypothesis that physical dynamics play a crucial role
in determining neural structure.