The robot setup mimicks the AIRO research setup with UR3e robots, Robotiq 2F85 grippers and Zed2i cameras.
- create a conda environment
conda env create -n robot-learning-playground
and activate it withconda activate robot-learning-playground
(use python3.9 for ikfast) - move into the
ur_sim
folder and runpip install -r requirements.txt
- manually install the
ur_ikfast
package using the instructions here (and test usingtest_ikfast.py
) - test that pybullet is working by running the
push_env.py
file. You should see the robot moving.
- activate the conda environment
- from the learning folder run
conda env update -f learning.yml