ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question
1

Moveit teleoperation of 3DOF arm with 3D mouse

asked 2018-06-12 10:03:03 -0500

blubbi321 gravatar image

In short, I want to control the endeffector of my 3DOF arm in cartesian space with a 3D mouse. (Very similar to https://answers.ros.org/question/2269... , only that my robot arm only has 3 DOF)

What I got working so far:

  • Moveit configuration generated
  • I can move the interactive marker, plan and execute paths in Rviz
  • I can move the arm into goal poses with orientation from python

Now my problem seems to be that due to the lack of DOFs I always need the orientation values for the EEF in the moveit interface to plan poses that I want my endeffector to end up in.

I attempted to

  • generate an IKFast solver using the Translation3D configuration/parameter => compiles, works, however still requires orientation?!
  • play with the tolerance as mentioned in https://bitbucket.org/traclabs/trac_i... (changed the value in the all_close() function)
  • tried to set position_only_ik: True in kinematics.yaml
  • took a look at https://github.com/ros-planning/movei... however I dont want any EEF rotation (also can set it), just an EEF x,y,z position

Now my question is: How can I get moveit to plan paths to EEF positions without needing to specify any orientation values?

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted
0

answered 2018-06-12 14:35:19 -0500

AndyZe gravatar image

updated 2018-06-12 14:37:26 -0500

If you have a decent background in linear algebra, it shouldn't be too hard to add this to the jog_arm package. First you would need to understand what the Jacobian is.

Otherwise, I'll get around to adding it eventually.

edit flag offensive delete link more

Comments

Ok, too bad .. thanks for your answer!

Then Ill leave the question open for now as there dont seem to be ROS "onboard" tools that are ready to be used.

Think Ill try for a hacky workaround and just add simulated fake joints at the endeffector so that the solver can find a solution there

blubbi321 gravatar image blubbi321  ( 2018-06-16 02:38:41 -0500 )edit

Question Tools

2 followers

Stats

Asked: 2018-06-12 10:03:03 -0500

Seen: 466 times

Last updated: Jun 12 '18