11.11.2009 - BIRS: Noise, Time Delay and Balance Control

Ian Loram vs Ramesh Balasubramaniam

• Passive versus active control of human balance
  • Passive control
    • passive joint meta-stabilization
      • Pin-joint model is not good for getting things to stand up
      • Joints conform during upright position.
    • muscle activity in standing balance
      • Muscles that appear to be on:
        • Soleus
        • Gastrocnemius
        • Ilipsoas
      • How much muscle activity is hidden from "sensing"?
        • VO2 = 20 W delta between standing and lying, 200 W delta between walking and standing
    • short range joint stiffness
      • Stiffness is high for small movements
      • Stiffness is low for large movements
    • control by intermittent activity
    • information flow in passive muscles
    • standing without passive mechanisms
  • Active control
    • The organization of task-specific solutions
      • PCA to determine dimensionality
      • UCM to separate task relevant vs task irrelevant
    • Does the cns understand physics?
    • Intermittency and prediction - Do they exist?

Manoj Srinivasan

• Optimal control can predict "forward leaning" based on asymmetric "muscles" doing co-contraction.

Jason Kutch

• Listening to actuators gives you information into how they act.
  • Signal dependent noise gives you information from the endpoint source about the muscles producing the force.
• Spike-triggered averaging
  • the average stimulus preceding a spike