i noticed, that grasp planning is very limited for the Katana, and i
found a description in the mailing list, that it is due to the fact,
that the Katana has only 5 DOFs. Since i only want to grasp a special,
smaller cylindric object, i would like to try to "plan the grasping"
simply by computing an IK-Solution for the toolcenter-point beeing in
the center of the object to grasp. I used Openrave with C++ for
computing trajectories for one joint-value configuration to another -
but now I want to use Python, and I can't find an example to compute a
joint-value configuration for the toolcenter-point matching a special 3D
point in space - is that possible?
Btw, i noticed the fastgrasping-example - there's some deprecated stuff
in it and I don't understand it exactly by skimming over, but is it
worth to try it - or does it work similar to the approach I want to try?
Richard Cubek, Dipl.-Ing.(FH)
University of Applied Sciences Ravensburg-Weingarten
Intelligent Mobile Robotics Laboratory
Phone: (0049) (0)751 501 9838
Mobile: (0049) (0)163 88 39 529
Sorry for asking about the fastgrasping example before trying it. Well,
it works nice, but I can't use it, because I need to stack/unstack
cylindric objects, and some grasp solutions for grasping the object on
the top of a stack would effect the object under it, too. And to my
second question: I guess, I have to iterate over different end-effector
transformations (suitable for grasping) and search for IK-Solutions.