See also:
http://cdr.stanford.edu/~baileys/WWWfiles/queue1.pdf
http://cdr.stanford.edu/~baileys/WWWfiles/bigpic1.pdf
 

Prioritized list of things to do

1 - Fix Sprawl (first since a one-shot item)
        - servo replaced
        - kickback diodes
        - PIC installation (sig. amt. of work)
        - buy and install parallel springs in legs

Rationale
        - prerequisite for #2

2 - Sprawl measurements
        - given leg angles and stride frequency
        - quantitatively measure velocities
        - qualitatively measure disturbance rejection

Rationale
        - See if we can get it to walk better
        - Try out some biomimetic ideas (footplacement, timing,
          equilibrium position regulation, variable stiffness torsional
          joints, mass distribution relative to hips)

3 - Explore limit cycles
        - test equilibrium finder for 2D model with given initial geometry
        - discuss limit cycle production thru state switching w/ Tomlin
        - look at cockroach trajectory and tripod switching

Rationale
        - See if there is a similarity between the cockroach body
          trajectory/leg duty cycles and stable variable state systems.
          (These are systems that maintain a limit cycle or converge to an
          equilibrium point by changing their state equations at different
          locations in the phase plane.  In terms of mobile systems, feet
          touch or lift off the ground, thus changing the state equations.
        - Want to see if we can design a legged system with compliance
          (or "complex stiffness") that will yield the same types of
          stable non-linear behavior

4 - Look closely at Full's mechanical properties results
        - characterize into a set of tradition components (springs,
          dampers, etc)
        - Can our plastics replicate these properties?
AND
4 - Build tiny SDM leg to test with Dudek/Meier setup at Berkeley

Rationale
        - The Dudek/Meier setup in Berkeley is experimental apparatus that
          allows them to make mechanical properties measurements on
          cockroach legs.  Outside of their community, the testing setup
          and results are not well understood.  At the least, a common
          frame of reference will be established (fundamental part of the
          MURI).
        - We also want to see how our engineering materials (ie plastics)
          behave in terms of mechanical properties.  Static evaluations
          are easy, but they seem to be doing more interesting dynamic
          evaluations.  In turn, this will lead to us being able to
          engineer the mechanical properties we want into our parts.
 

5 - Install big servos on sprawl, torsionally compliant hips
        - vary stiffness of torsional springs

Rationale
        - Almost a detail issue under #1, but heavier time commitment.
        - Also, it would be nice if the compliant part of the hip was an
          SDM component (plastic) rather than tradition spring and
          bearings....

6 - Tuned/Variable compliance hip in SDM

Rationale
        - Way down the road at this point - we need to have a better idea
          of what we want and can achieve (which should be facilitated by
          the previous items

6 - Complete 2D model with event checker
        - can deal with foot contact, switching tripods
        - matlab's checker is not robust enough - too much time wasted
        - make our own integrator

Rationale
        - test bed for ideas about mechanical properties and
          configurations
        - Useful if it takes less time than actually building the device

7 - Mini-Sprawl - small version of the big Sprawl
        - mouse-sized version of Sprawl built in SDM

Rationale
        - Identify issues with making very small and complex SDM'ed parts