sled-TIAGo-learning

Roadmap

1. Study ROS Tutorial
2. Study TIAGo Tutorial
3. Read TIAGo Handbook
4. Try out examples in this repo

Env Setup

1. Install ROS-noetic (already done on whistler), then

source /opt/ros/noetic/setup.bash

2. Install TIAGo package, then

source ~/tiago_public_ws/devel/setup.bash

3. Install more packages (already done on whistler /usr/bin/python)

git submodule update --init
cd third_party
pip install tracikpy/  # check https://github.com/mjd3/tracikpy for dependencies
cd Grounded-Segment-Anything/GroundingDINO && python setup.py build && python setup.py install # set CUDA_HOME
cd ../../
pip install -e Grounded-Segment-Anything/segment_anything/

Install itself

cd <project_dir>
pip install -e .

Run examples

python examples/xxxxx

Caveat

1. Never run on real robot before verifying in Gazebo.
2. Use Ctrl+C to close Gazebo, be patient
3. Never push the "⏻" button to shutdown TIAGo, use the button with green light.

Hints:

1. set settings.json in .vscode for better python file parsing

{
    "files.associations": {
        "*.tcc": "cpp",
        "ratio": "cpp",
        "condition_variable": "cpp"
    },
    "python.analysis.extraPaths":[
        "/opt/ros/noetic/lib/python3/dist-packages",
        "/path_to_your_installed/tiago_public_ws/devel/lib/python3/dist-packages"
    ],
    "ros.distro": "noetic",
    "python.autoComplete.extraPaths": [
        "/opt/ros/noetic/lib/python3/dist-packages",
        "/path_to_your_installed/tiago_public_ws/devel/lib/python3/dist-packages"
    ]
}

2. Every time run

source /opt/ros/noetic/setup.bash
source ~/tiago_public_ws/devel/setup.bash

TODO

1. check gripper open position for real robot