-- SeanBailey? - 24 Jun 2003

Biologically Inspired Adaptation

I am only planning on doing one or two of these, but this is a good place to think about things


Time of Response

  • Tap into a high EMG frequency muscle
  • Measure the time of response to perturbation
  • Same as old kugle setup, with similar directions and data acquisition

  • Tells you
    • Lower limit of the time of response (so what?)

  • Tasks
    • Locate high frequency muscle
    • Use current setup and code
      • apply perturbation at time interval from initiation of EMG
      • measure time of response (change from normal)


Reaction to cyclic perturbations

  • Instead of applying perturbations in the vertical direction, apply them fore-aft (or maybe lateral)
  • While excursions are about the same in the 3 directions, fore-aft power expenditure (and thus the changes in energy) are much greater in this direction
  • This is also the main thing that changes when going from flat ground to vertical climbing
  • Tells you
    • If perturbations in this direction are attended to
    • This may be a possible measure of how good the locomotion is
  • Accelerometer on robot body (and in the foot)

  • Tasks
    • Work out what adding a different frequency sinusoid looks like
    • Try again in the vertical direction
      • verify previous results?
    • Fore-aft direction
      • Verify
        • Excursions
        • Energy/power expenditure
        • Mathematics
    • Build new actuator/sensor
      • Laptop HDD
      • Laser sensor for position
      • Force sensor
        • Semiconductor straingage - design beam and flexures
        • Mouse pointer - sensitivity?
      • Electronics and interface to operate
      • Calibrate with known masses and compliances
    • New mounting design
      • prismatic slider
      • flexures
      • direct (like current)
    • Put accelerometer on back of
      • Animal
        • Verify flat running force patterns
        • New data on vertical or near vertical running
        • Need thin wire (order from company)
        • Need more animals!
      • Robot
        • See if there is a correlation between actually moving and force patterns
        • Ask Jonathan about fore-aft ground reaction forces
    • Voice coil on the back of a free-running animal
      • Use with EMGs for descending frequency
      • What is the lower limit of frequencies for a voice coil (need 10-25 Hz)
      • How much mass is there that it moves around?
    • Climbing
      • Simulate by using air jets
      • Relocate bug to being at an angle from the kugle


Ground reaction forces and sprawl angles up and across slopes

  • The gecko seems to change strategy from flat ground to climbing, does the cockroach as well?
  • Have the animal run up (and across slopes) with force sensors underneath and high speed video
  • Do the same thing with an accelerometer glued to the back of the animal (whole body only)
  • Tells us
    • Are insects’ strategies different than Geckos? (do they change from pushing outwards to pushing inwards?)
    • Ground reaction forces for different slopes and angles
    • Change in sprawled posture
    • Change in foot orientations
    • Overall variations in accelerations of the body

  • Tasks
    • Have others done this work? Kellar? Others? Read Dan's paper... Look up vertical running
    • Ground reaction forces
      • Figure out the accelerations and the forces we expect to see
      • Research types of force sensors
        • Mouse pointer (piezo? Electronics, sensitivity)
        • Bought sensor
        • Homemade (with strain gages) – structure? flexures?
        • Full lab type - Heglund 1981
      • Whole body versus individual leg GRFs
      • Accelerometer on back? Gives whole body forces?
    • Sprawl angles
      • high speed video with a marker at foot and on body
    • Basic track design
      • Basic U channel
      • Flat plate (with bright light behind, or vaseline around)
      • Squashed cylinder (ala CGW)
      • What slopes and materials can the two types of cockroaches run up?


Convex and concave surfaces

  • Have the animal run over convex/concave surfaces with embedded force sensors - does the strategy that insects adopt change with the surface they are running over?
  • High speed video to look for changes in posture
  • What path does it choose when given a corrugated roof surface?
  • Tells us
    • If strategy is dependant on surface curvature

  • Tasks
    • Run animal up corrugated surface - does it pick convex or concave?
    • Same as above

 
This site is powered by the TWiki collaboration platformCopyright &© by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback