search

borglab / GTDynamics

Full kinodynamics constraints for arbitrary robot configurations with factor graphs.
BSD 2-Clause "Simplified" License
39 stars 10 forks source link

Add models from the DARPA SubT challenge #289

Closed varunagrawal closed 2 years ago

varunagrawal commented 2 years ago

Fixes #281.

varunagrawal commented 2 years ago

ping @gchenfc

gchenfc commented 2 years ago

Sorry I must have missed this the first time around. I think the target branch is wrong? Or if you're intentionally trying to keep a separate branch, maybe you can create a separate PR first to merge master (or just perform locally) before this PR so it's easier to read?

gchenfc commented 2 years ago

Oh nevermind, but how come the "files changed" diffs are all messed up on github?

varunagrawal commented 2 years ago

Yeah that's because I merged in master and Github doesn't seem to like that. It should resolve itself once we merge in robot-models.

varunagrawal commented 2 years ago

I guess I can remove the inner subT folder.

varunagrawal commented 2 years ago

This is fine, but perhaps in the future we should consider also reorganizing the urdfs / sdfs into robot-oriented directories (e.g. subt / spider / a1 / cablerobot / etc) instead of sdf / urdf / subt. Since loading syntax doesn't change depending on the file type, and it doesn't really make sense to have folders for file types and a folder that's categorically named that contains sdfs.

The calling syntax is different. And if we have a URDF of a robot, why would we need an SDF and vice-versa? This is why I felt having a robot-centric naming convention was not helpful since the directories would just have the one file and unnecessarily add more typing to load the file aka why should a user type cablerobot/cablerobot.urdf when they can simply do kUrdfPath/cablerobot.urdf?

Also just realized, why are most of these files except cerberus... in a subt/subT subfolder?

Yeah I don't know what I was thinking (or not thinking). Updated this in the last commit, thanks!