You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
problem (probably pr2 hardware or pr1040 specific issue)
(This issue is using version after #63) rosrun mohou_ros reset_to_home.py on real robot
causes the following error
[ERROR] [1662316480.793608]: 1 non-controlling joint angles are not consistent with home position.
{'joint_name': 'head_tilt_joint', 'angle': 1.1361046232931888, 'angle_desired': 1.0728538912009142, 'error': 0.06325073209227461, 'threshold': 0.03490658503988659}
However, in the gazebo simulator, after roslaunch pr2_gazebo pr2_empty_world.launch the same command works without problem.
Also, I conducted additional experiments. running the same command changing the initial tilt angle.
case1: tilt_joint is set to the highest value (neck is pointing to the ground)
case2: tilt_jotin is set to the lowest value (neck is pointing to the sky)
In case2, the command works well, but in case1 the same error occurred.
Also I check the /head_traj_controller/follow_joint_trajectory/feedback topic when testing on case1. And the output is like the following. The output shows that desired neck position is set correctly, but controller cannot drive the joint to the desired angle. So, I conclude that it is a hardware-rooted problem rather than a scikit-robot software issue.
Actually, the same problem occurs when using euslisp. The following code compute the difference between desired and actual joint angle. and output was 3.26 [deg].
problem (probably pr2 hardware or pr1040 specific issue)
(This issue is using version after #63)
rosrun mohou_ros reset_to_home.py
on real robotcauses the following error
However, in the gazebo simulator, after
roslaunch pr2_gazebo pr2_empty_world.launch
the same command works without problem.Also, I conducted additional experiments. running the same command changing the initial tilt angle.
case1: tilt_joint is set to the highest value (neck is pointing to the ground)
case2: tilt_jotin is set to the lowest value (neck is pointing to the sky)
In case2, the command works well, but in case1 the same error occurred.
Also I check the
/head_traj_controller/follow_joint_trajectory/feedback
topic when testing on case1. And the output is like the following. The output shows that desired neck position is set correctly, but controller cannot drive the joint to the desired angle. So, I conclude that it is a hardware-rooted problem rather than a scikit-robot software issue.Actually, the same problem occurs when using euslisp. The following code compute the difference between desired and actual joint angle. and output was
3.26
[deg].home position
home_position.yaml
The text was updated successfully, but these errors were encountered: