Description of Cartesian Interface from documentation:
The YARP Cartesian Interface has been conceived in order to enable the control of the arms and the legs of the robot directly in the operational space, that is instead of commading new configuration for the joints, user can require for instance the arm to reach a specified pose in the cartesian space, expressed as a combination of a 3d point to be attained by the end-effector (the center of the palm in this case) along with the desired orientation given in axis-angle representation.
It took me quite a lot of time to get the Cartesian Interface working, because of all the dependencies and configurations which are needed to be accomplished before using this interface.
- I plan to use this interface in some experiments with iCub, because it provides an easy way how to control robot arms. So wouldn’t have to focus on image processing and motor control a lot, therefore I could focus more at application of reinforcement learning to learn iCub reach/grasp objects.
- Later, I will do all that stuff on my own, without the use of Cartesian Controller.
If you also have issues with this interface have a look at “cartesiancontroller issues” forum thread. Or contact me and I would help you.
Some screenshots from my experiments…
June 2018 M T W T F S S « Mar 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
iCub on Twitter