Thanks for your reply！This week I’ve also had a try for the Articulation interface to achieve the same task above, just with the guide you provided and Franka’s FollowTarget task as my reference.
It seems the action is successfully applied to the arm of Fetch, and according to the output of compute_end_effector_pose , the end effector seems to be right at the target position, but actually there’s a always a deviation between the actual position of end effector and my target position. I suspect the deviation is caused by the incorrect robot base pose, but I don’t know how to fix that. By the way, the same implementation works as expected for Franka, the deviation only exists on Fetch Robot.
Could you please have a look at the problem in that post?