airo-models
Curated URDFs and 3D models of the robots and gripper used at airo.
Installation
airo-models is available on PyPi and can be installed with pip:
pip install airo-models
Usage
Example of loading a URDF from airo-models, customizing it and writing it to a temporary file:
import airo_models
robotiq_urdf_path = airo_models.get_urdf_path("robotiq_2f_85")
robotiq_urdf = airo_models.urdf.read_urdf(robotiq_urdf_path)
airo_models.urdf.replace_value(robotiq_urdf, "@type", "revolute", "fixed")
airo_models.urdf.delete_key(robotiq_urdf, "mimic")
airo_models.urdf.delete_key(robotiq_urdf, "transmission")
robotiq_static_urdf_path = airo_models.urdf.write_urdf_to_tempfile(
robotiq_urdf, robotiq_urdf_path, prefix="robotiq_2f_85_static_"
)
To check which models are available:
from airo_models.files import AIRO_MODEL_NAMES
print(AIRO_MODEL_NAMES)
>>> ['ur3e', 'ur5e', 'robotiq_2f_85']
Modeling conventions
The standard convention we follow is X+ forward, Z+ up.
For cameras, we follow Z+ forward through the eye of the camera, X+ right. The origin of the camera is at the center of the (left) lens. Left is defined egocentric of the camera (i.e. looking out of the eyes of the camera).
For grippers, we follow Z+ pointing outwards from the fingers and X in the direction in which the parallel gripper closes its fingers. The origin of the gripper (base_link) is at the mounting point of its base.
Development
Local installation
- Clone this repo
- Install the dependencies using
uv sync
- Initialize the pre-commit hooks
uv run pre-commit install
- Run the tests with
uv run pytest .
Releasing
Releasing to PyPi is done automatically by github actions when a new tag is pushed to the main branch.
- Update the version in
pyproject.toml.
git add pyproject.toml
git commit -m ""
git push
git tag -a v0.1.0 -m "airo-models v0.1.0"
git push origin v0.1.0
This was set up following this guide first and then this guide.