Lightweight 2D ROS 2 Simulation & Custom RViz UI

📖 About The Project

This project provides a minimalist, fast, and mathematically precise 2D simulation for debugging autonomous navigation algorithms without the overhead of heavy 3D engines.

The environment is fully containerized via Docker. It features a custom RViz2 visual interface that publishes complex geometric markers (hexagonal grids, background overlays) while successfully addressing common 3D rendering issues like Z-fighting.

✨ Key Features

• 🐳 Dockerized Environment: Zero setup required. Flatland and all dependencies are pre-installed in the provided Docker container based on the osrf/ros:humble-desktop-full image.
• 🚀 KISS Architecture: Lightweight physics powered by the Flatland 2D simulator (Box2D). Runs smoothly even on low-end hardware or CI/CD pipelines without GPUs.
• 🤖 Gazebo-Matched TB3 World & Kinematics: Accurately replicates the classic 3D TurtleBot 3 world from Gazebo into a fast 2D environment. It maintains the precise physical dimensions of the TB3 Waffle (281 x 306 mm) and the exact 6.4 cm offset between the base center and the drive axle, ensuring highly realistic navigation and collision behaviors.
• 🧩 Collision Layers Solved: Correctly configured physics layers (static and robot). The robot's LiDAR freely passes through its own footprint but reflects off walls.
• 🎨 Custom RViz2 UI: A ready-to-use Python module (marker_array_publish.py) for batch publishing MarkerArray messages (hexagons, circles, squares) with perfectly calculated Z-order.
• ⏱️ Robust Bringup: An automated launch sequence that handles delays for the Flatland server, TF transformations, and asynchronous robot spawning.

---

⚙️ Installation & Build

Prerequisites:

• Docker installed on your host machine.

1. Create an isolated workspace for ROS2 and clone the repository:

mkdir -p ~/your_folder/src
cd ~/your_folder/src
git clone https://github.com/rprakapovich/tb3_flatland.git

2. Navigate to the docker directory and build the container:

cd tb3_flatland/docker
docker build -t flatland_sim .

🚀 Usage

1. Run the Docker Container To run the container with graphical interface support (X11 forwarding for RViz2), use the following command:

cd ~/your_folder/src/tb3_flatland/docker
bash launch_docker.bash

2. Project building:

colcon build --packages-select tb3_flatland --symlink-install
source install/setup.bash

3. Launch the Simulation Stack Once inside the container, launch the simulation, RViz2, and the custom UI markers with a single command:

ros2 launch tb3_flatland tb3_flatland_bringup.launch.py

4. Move the Robot Open a new terminal, attach to the running Docker container, and start the steering GUI to control the robot:

# In a new host terminal:
docker exec -it flatland bash

# Run the steering GUI
ros2 run rqt_robot_steering rqt_robot_steering

Note: The robot listens to the default /cmd_vel topic.

🏗 Architecture Details

Collision Layers Trick

This project solves the common 2D simulator issue where a robot's LiDAR detects its own body as an obstacle.

• The robot resides on the [robot, static] layers.

• World walls reside on the [static] layer.

• The Laser plugin is configured to scan only the [static] layer while explicitly ignoring the base_link body it is attached to.

Custom RViz Marker UI

The marker_array_publish.py node programmatically generates a hexagonal grid layout using Marker.TRIANGLE_LIST. It applies specific Z-axis offsets (z = -0.1 + marker_id/100) to prevent Z-fighting and Alpha-sorting rendering artifacts in the Ogre3D engine used by RViz2.

🗺 Roadmap

• Configure base physics and collision layers in Flatland.

• Build custom marker tooling for RViz2 (hex-grids).

• Containerize the workspace with Docker.

• Add Multi-Robot support (spawning multiple agents with correct namespaces and TF trees).

• Integrate with Nav2 (Navigation 2 stack).

🤝 Contributing

Contributions and Pull Requests are welcome! If you prefer simple, elegant, and lightweight robotics over overly complex setups, feel free to contribute.