Hi David, (1) This is difficult to debug without more information. One thing you could do is display the pose of the robot corresponding to the last collision in rviz. This tutorial - http://www.ros.org/wiki/motion_planning_environment/Tutorials/Tutorial%20B - helps you do that. (2) Also, using the rviz "select" button, you can click on the markers and they will give you more information on what is colliding. (3) Are you moving the base of the robot around? We are working towards a release that handles motion of the robot base properly - the current collision environment is not yet fully tested for motion of the base. Regards, Sachin On Thu, Aug 5, 2010 at 7:24 PM, David Lu!! wrote: > Hey all, > As I mentioned in another thread, I'm having a problem with collision > detection. I want to test to see if the robot is colliding with itself. I've > loaded the URDF into robot_description, I'm publishing joint_states, I've > set default_collision_operations, and defined groups. Here's how I'm > launching planning_environment. > name="environment_server_myrobot"> >    > > > > > > > >          > My robot generously fits into a 2 meter cube. However, when I run collision > detection (i.e. call GetStateValidity), the collisions the algorithm finds > are far outside the 2 meter cube. Here's one marker: > --- > header: >   seq: 150 >   stamp: >     secs: 1281057846 >     nsecs: 457145271 >   frame_id: BASE_FRAME > ns: PIECE1+PIECE2 > id: 150 > type: 2 > action: 0 > pose: >   position: >     x: -67.8300476074 >     y: -29.5642662048 >     z: 11.4108705521 >   orientation: >     x: 0.0 >     y: 0.0 >     z: 0.0 >     w: 1.0 > scale: >   x: 0.01 >   y: 0.01 >   z: 0.01 > color: >   r: 1.0 >   g: 0.800000011921 >   b: 0.0399999991059 >   a: 0.600000023842 > lifetime: >   secs: 0 >   nsecs: 0 > frame_locked: False > points: [] > colors: [] > text: '' > mesh_resource: '' > mesh_use_embedded_materials: False > Why would there be a collision all the way out there? My only thought is > that all of my meshes are scaled to a factor of .01, but even manually > scaling the markers down by that much, they don't correspond with the space > that the robot occupies. > Any thoughts? > -David!! > _______________________________________________ > ros-users mailing list > ros-users@code.ros.org > https://code.ros.org/mailman/listinfo/ros-users > > -- Sachin Chitta Research Scientist Willow Garage