[ros-users] What is the best way to develop Gazebo models?

[John Hsu]

Hi Patrick,

For model editing in gazebo, we generally create models using the following
steps:

1) Make sure your model is static - or start the simulator paused
2) Import all your bodies and geometries. Make sure the bounding boxes
match your visualization using View->Bounding Boxes
3) Layout all the links in the model so that their poses are correct
4) Create the joints that connect each body in the body.

It's key to make small changes and then test in gazebo.

In terms of visualization capability in gazebo, in summary, gazebo currently
can visualize following entities:

   - visual mesh
   - bounding box for the collision mesh
   - coordinate axis for the inertial body, or Center of Mass location
   ("Show Physics")

It cannot visualize the following items:

   - collision mesh
   - coordinate axis for the collision mesh
   - coordinate axis for the visual mesh
   - coordinate axis for the model (canonical body)

But it is on our todo list to update the GUI to make model building more
convenient.  Please feel free to submit patches or open feature request
tickets.

If you are developing models in URDF, David Lu!! has put together a nice set
of tools to do so, please see his stack:

http://www.ros.org/wiki/urdf_tools

Hope this helps.
John

On Wed, Oct 13, 2010 at 6:48 PM, Patrick Doyle <wpdster at gmail.com> wrote:

> Can you folks give me tips on developing my own Gazebo models?  Is
> there an easy way to visualize how the different coordinate systems in
> a link (joint, inertial, visual, and collision) interact with each
> other?
>
> --wpd
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osuosl.org/pipermail/ros-users/attachments/20101214/e7d2ac1b/attachment-0002.html>