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

Revision history [back]

click to hide/show revision 1
initial version

Your robot arm is a mechanical system, and it may not be capable of doing better. You can not set an arbitrarily small number in software and expect that the robot is able to achieve it.

That said, the positioning of the arm links that you choose to use for the grasp can make this better or worse. The further you make the End Effector reach out horizontally from the base, the worse this metric usually gets (because you are forcing the joint motors to actively work to fight gravity.)

Your robot arm is a mechanical system, and it may not be capable of doing better. You can not set an arbitrarily small number in software and expect that the robot is able to achieve it.

That said, the positioning of the arm links that you choose to use for the grasp can make this better or worse. The further you make the End Effector reach out horizontally from the base, the worse this metric usually gets (because you are forcing the joint motors to actively work to fight gravity.)

If you have evidence this is a ros software issue, it could be any number of things. To find the cause, you need to run some tests to narrow down the issue:

  • If you command a joint to a specific position, how much error does it have after it stops moving?
  • Are the joint positions being read-back accurately?
  • Is there a problem with the timing of when the joints are told to stop moving (too early or too late)?
  • is there a problem with the Inverse Kinematics (IK) calculation (e.g. an incorrect dimension in the robot model)?