sdf_to_urdf.md

Authors - Alexion Ramos & Ilan Hascal

Creating a Gazebo Model

For some users it is easier to create a model through the given Gazebo simulator in the Model Editor than having to write a custom file. The problem that arises is using objects and specific services else where in a project can be difficult due to the extension names of the file.

Building Saving a model

When you want to build a model in Gazebo it is best that you do it in the Model Editor that you can get you with CTRl m

Once the model is built and you save it to the right directory it is saved as an .sdf file

Why is .sdf not useful for Autonomous Pacman and how it conflicts with Services?

For the sake of the autonomous pacman project, our goal was to implement collectibles that pacman could pick up to make him "invincible". Though ,the gazebo SpawnModel & DeleteModel services expect a .urdf file in order to track the model down to Spawn or Delete. Example of spawning a custom .urdf is below.

    spawn_model_client = rospy.ServiceProxy('/gazebo/spawn_sdf_model', SpawnModel)
    spawn_model_client(
    model_name='TESTCYLINDER',
    model_xml=open('/my_ros_data/catkin_ws/src/ros-autonomous-pacman/models/TESTCYLINDER/TESTCYLINDER.urdf', 'r').read(), 
    initial_pose=Pose(),
    reference_frame='world'
    )

Documetnation for Spawn - http://docs.ros.org/en/jade/api/gazebo_msgs/html/srv/SpawnModel.html Documentation for Delete - http://docs.ros.org/en/jade/api/gazebo_msgs/html/srv/DeleteModel.html

Solution

After reasearching it was with great pleasure that we found an open source library that allows you to put input the name of the .sdf file and then converts it to .urdf as output to where you specify. It is a really straight forward process that can be done by using the following links below

Github repo - https://github.com/andreasBihlmaier/pysdf Youtube video instruction - https://www.youtube.com/watch?v=8g5nMxhi_Pw