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

Robot model collapsing in Gazebo

asked 2016-05-19 09:46:10 -0500

Laurens Verhulst gravatar image


I am using ROS indigo in combination with the included Gazebo version. I am trying to control a robot arm in combination with ROS control in Gazebo. The robot arm I am controlling is the Kinova Mico arm. I am using the URDF model from the WPI Jaco package, however I needed to modify it slightly to be able to have the robot use the correct inertias for each link.

When I spawn the arm in a paused physics simulation, the arm spawns with all the joints in the correct position. However, when the simulation is unpaused, the arm collapses and all links jump to the origin of the world. I have a figure of this, however cannot upload it due to the lack of points. For the same reason, I cannot upload the URDF model.

I found a possible solution that said that the inertias are too low. The inertias for the different links are however in the order of 10e-4 so should be fine. At least, I believe the URDF model asks for inertia in kg*m2.

If any of you have experience with this or know a possible solution, that would help be out greatly. Thanks.

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted

answered 2016-05-20 01:38:05 -0500

I think my answer here covers the most common knobs you can try tweaking. If you're spawning the robot without controllers, the combination of inertial parameters of the robot and physics engine parameters appear to be to blame.

edit flag offensive delete link more


Thanks for your reply! I will try this out and post the results here.

Laurens Verhulst gravatar image Laurens Verhulst  ( 2016-05-20 02:39:24 -0500 )edit

I have located the issue, it is because 2 of the 6 links are weirdly shaped. This causes the simulation to be instable. I am still trying to fix this issue.

Laurens Verhulst gravatar image Laurens Verhulst  ( 2016-05-23 01:43:22 -0500 )edit

Ok, I fixed the issue, it was indeed a combination of inertia and mass of two specific links.

Laurens Verhulst gravatar image Laurens Verhulst  ( 2016-05-23 06:34:35 -0500 )edit

Question Tools



Asked: 2016-05-19 09:46:10 -0500

Seen: 2,458 times

Last updated: May 20 '16