🚀 go-pugleaf

Hi,

I have a few questions I would like ask this hopefully well-informed group.
I'm new to the field of robotics applications--my background is in
structural dynamic interactions with control systems.

I am interested in improving the *dynamic performance* of a 7 d.o.f. spherical
manipulator (with a reach of approx. 1.5 m) by compensating for the effects
of joint flexibility (due largely to the harmonic drives).


TOPIC 1: Defining and Measuring Dexterous Manipulator PERFORMANCE

 The performance measures I am most interested in are an external measure of
 end-effector position and orientation (i.e. "pose") accuracy, stability, and
 jitter (where I define accuracy to be the error in the desired "pose" and
 the mean "pose", stability to be the RMS error in the "pose" about its mean,
 and jitter to be the RMS error in the "pose" velocity).  Repeatability of
 the "pose", as well as applied force/torque accuracy, stability, and jitter
 will eventually be considered.
 
Q1 - Are these reasonable performance measures for a supposedly dexterous 
         manipulator?  Did I forget something important?  Am I clueless?

Q2 - What off-the-shelf systems have people used (and are happy with) to
         externally measure the static and dynamic end-effector "pose"
         performance?


TOPIC 2:  Compensating for JOINT FLEXIBILITY

 If you are aware of any previous or current work in this area please let me
 know.  Article references, theses, names, whatever...


Thanks for your feed-back.  I will post a summary of the helpful responses
under the headings `Manipulator Performance Measure' and `Joint Flexibility'

Dan

------------------------------------------------------------------
 Daniel A. Rey                                                          
Internet: dan@sputnik.mit.edu
 MIT Space Engineering Research Center
 ------------------------------------------------------------------