Robot Workspace (robot-ws)

This is the primary ROS 2 (Humble) workspace governing the low-level nodes for the SBC (Single Board Computer) deployed on the robot (e.g., Radxa Zero).

It encapsulates robust sensor interfacing, footprint publication, inverse/forward kinematics (odometry), motor driving, and advanced reinforcement-learning driven stabilization policies.

🏗️ Structure

Included submodules orchestrate various components of the robot:

• ddsm115_driver: Interfaces with the DDSM115 direct-drive motors.
• imu_driver: Connects and publishes raw IMU readings.
• odometry_node: Fuses sensor logic into robust kinematics data.
• robot_tf_publisher: Coordinates standard static/dynamic tf2 transforms and URDF definitions.
• robot_launch: The central launch nexus encompassing logic to boot single or composed modular sets of nodes.
• stabilization_controller: Real-time balancing proxy utilizing ONNX-runtime evaluated AI models for stabilization.
• ydlidar_ros2_driver: Customized code tailored for the TMini Laser Scanner (replaces the standard YDLidar node).

🚀 Compilation & Development

You can natively compile the system on a matching aarch64 / x86_64 machine or utilize the supplied Docker configuration.

Using Docker (Highly Recommended)

We provide a simple Dockerization technique to ensure compiling code always perfectly mirrors the robot's ecosystem environment.

1. Ensure that all modules are downloaded and up to date

git submodule update --init --recursive

2. Spin up the Docker Container: We recommend using docker compose to effortlessly mount this workspace and boot into it:

   docker compose run --rm robot-dev
   # Alternatively via run: docker run --rm -it -v $(pwd):/robot-ws -w /robot-ws ros:humble-ros-base bash

3. Compile the Workspace: Inside the Docker container terminal (which automatically sources the base ROS distribution), simply invoke colcon:

   colcon build --symlink-install

🎮 Launching the Robot

Always source your newly compiled workspace configurations before attempting a launch:

source install/setup.bash

To initialize the primary process grouping, utilize the core components residing in robot_launch:

ros2 launch robot_launch robot.launch.py

Launch Arguments

You can easily pass CLI arguments to tailor behavior:

• enable_motors (Default: true) — Boots up motor communications via DDSM115 Node and associated Odometry Node.
• enable_rl_balancer (Default: false) — Unlocks the AI balancing mechanics utilizing the ONNX runtime module.

Example:

ros2 launch robot_launch robot.launch.py enable_rl_balancer:=true

Dev UI

Now ou can visualize ROS2 data using the Foxglove UI. You need to log in with your Google account at the following link:

https://foxglove.dev

after that select --> ROSS2, and paste the web socket adress from running foxglove_bridge instance that is lives inside the ross2 as package.

By default it's:

ws://127.0.0.1:8765

when you connected succsesefully, import the foxglove panel file which is will be provided soon