Gurdy URDF purpose

I am failing to understand the exercise to build Gurdy URDF. The exercise asks us to create a Gurdy robot with all the information already given. I don’t see how it makes me think how to build the URDF because the meshes for the links have their own origin which is unknown and I simply hit and trial until I get the right values to place it. I am failing to understand how this would help me build a URDF.

The tree for URDF was given and if one tries to follow the tree without the meshes its not possible to understand why the joints are placed in the ways given.

URDF’s are supposed to be built with a proper plan and information. Of course we would have to make minor tweaks which is most one should do.

For example:
I would assume the basic geometry given for collisions to have same origin as the child link joint and give values accordingly. But when I import the meshes its way off than expected.

The only thing that is making me think on would be the python script which I don’t believe the course is designed for .

I honestly felt like I spent too much time on the gurdy urdf without direction until I figured out the correct values and it wasted my time.
Can anyone help understand the unit?

Hi,

I can imagine it must have been difficult for you. It’s true that the unit probably needs some extra info about that. But it’s designed to be like that to show you that creating models in URDF is not always as simple as the examples done earlier, where everything was centered and placed nicely for you. By experience, most 3D models are decentered, don’t have the origin in the joints, don’t render properly, the collision meshes don’t match or even don’t exist or use 3D primitives due to system performance.

In the exercise, the distances provided, as you well have done, are the minimum to get the exercise working. Visualizing the 3D meshes in blender can help you see the origin of the meshes. It’s also a good practice to build the model piece by piece, validating and visualizing the link frames, the center of masses, and so on. It helo a lot to visualize how it will work.

But for sure, we will take your feedback into account at the next iteration of the course we will include or even create new units to talk and deapen about this.