[ros-users] Problem with PR2 Simulator IK Tutorial

Sachin Chitta sachinc at willowgarage.com
Wed Mar 10 01:02:57 UTC 2010


Hi Gal,

That tutorial is outdated. Sorry about that. You can have a look at more
recent tutorials for working with the ik at:
http://www.ros.org/wiki/kinematics_pr2/Tutorials

We are in the process of releasing the kinematics code with a new ROS api.
The api changes include an update to the name of the stack, so the tutorials
will move to
http://www.ros.org/wiki/pr2_kinematics/Tutorials

Expect the api changes to happen within the next few days. I will send an
email to the ros-users list with more information when we are ready and
released.

Regards,
Sachin



On Tue, Mar 9, 2010 at 4:39 PM, Gal Peleg <galpeleg13 at gmail.com> wrote:

> Hi guys, I hope all is well.
>
> My classmate and I are relatively new to ROS.  In our attempts to run
> through the "Moving the arm through a Cartesian pose trajectory using
> inverse kinematics <*
> http://www.ros.org/wiki/pr2/Tutorials/Moving%20the%20arm%20through%20a%20Cartesian%20pose%20trajectory?action=fullsearch&context=180&value=linkto%3A%22pr2/Tutorials/Moving+the+arm+through+a+Cartesian+pose+trajectory%22
> *> " tutorial, we ran into some difficulties.
>
> We managed to get all the files/packages to compile as outlined in the
> tutorial, but were unable to execute the last step. In particular, the
> turoial asks you to execute:
> python scripts/test_ik_trajectory_tutorial.py
>  after running:
> bin/ik_trajectory_tutorial
> which fails.  The script says that the trajectory failed, but the actual
> reason for failure appears to be a segmentation fault due to a null pointer
> in ik_trajectory_tutorial.cpp after the line:
>      bool ik_service_call = ik_client.call(ik_request,ik_response);
> when the code tries to access the data stored in ik_response.
>       solution[i] = ik_response.solution.joint_state.position[i];
>
> The addresses appear to be valid, but actually accessing any value within
> position leads to the segfault.  We attempted to trace back the calls
> through pr2_arm_ik_node, pr2_arm_ik_solver, and pr2_ik_arm, but were unable
> to determine exactly what went wrong when trying to convert the Cartesian
> coordinates in joint space.
>
> Any help is appreciated. Really, all we are trying to do at the end of the
> day is become familiar with a working implementation of an IK solver for the
> PR2 arms. So if anyone knows of a better (working) tutorial, that would be
> great as well.
>
> Thanks again!
>
> Gal & Sam
>
>
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>
>


-- 
Sachin Chitta
Research Scientist
Willow Garage
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ros.org/pipermail/ros-users/attachments/20100309/e61ad8fe/attachment-0004.html>


More information about the ros-users mailing list