Somatic Robots: Neural-Structural Integration - Time Change

Somatic Robotics approaches the design of physical structure and control intelligence as a single integrated task.  The Mind / Body duality of western thought, and the associated separation of control theory from mechanical engineering, are inadequate at providing insight to biology or future robotic systems.  To understand how we control motion, we need to understand the physical mechanism being moved. Emerging theories of vertebrate physiology are overturning the traditional bone-centric model of the body in favor of a "tensegrity" ("tension" + "integrity") model, in which the primary load paths are in the continuous tension network of the fascia (connective tissue and muscles). In turn, research and development at NASA Ames into dynamic tensegrity robots shows how these "soft machines" may be controlled through neuroscience inspired methods based on decentralized rhythmic controllers called "Central Pattern Generators" (CPG's).   The key insight is that both the structure and the controls share similar qualities of compliance, modularity, and the ability to synchronize into a unified neural-structural dynamical system.   This novel architecture has the potential to revolutionize future bio-inspired robots and provide new approaches to planetary exploration.

Speaker: Vytas SunSprial, NASA

Room 381T

Editor's Note: The time and location for this event have changed from later in the day.