ROS2 Humble NVIDIA Docker Template

A Docker setup for ROS2 Humble with NVIDIA GPU support, Gazebo, RViz2, and TurtleBot3. The goal is to get a working ROS2 environment running with minimal effort.

What's Inside

• ROS2 Humble Hawksbill
• NVIDIA GPU support via Container Toolkit
• Gazebo simulation
• RViz2
• TurtleBot3 (Burger)
• A basic publisher/subscriber package (pubsub_pkg) to test things out
• Mounted workspace so edits on your host show up in the container immediately

Requirements

You need these installed on your machine before starting:

• Docker
• Docker Compose
• NVIDIA GPU drivers
• NVIDIA Container Toolkit
• Linux with a desktop (needed for Gazebo and RViz2)

Renaming the Project

When using this template on GitHub, just name your repo whatever makes sense for your project. After cloning, if you want to rename the container too, open docker-compose.yml and change the container_name and image name fields.

Setup

1. Use this template on GitHub

Click "Use this template" and create your own repo.

2. Clone it

git clone https://github.com/YOUR_USERNAME/YOUR_REPO_NAME.git
cd YOUR_REPO_NAME

3. Allow GUI apps (do this after every login)

xhost +local:docker

4. Build the image

docker compose build

5. Start the container

docker compose up -d

6. Enter the container

docker exec -it ros2_ws_container bash

Build the Workspace

Inside the container:

cd /ros2_ws
colcon build
source install/setup.bash

Run the Example Nodes

Terminal 1:

ros2 run pubsub_pkg publisher

Terminal 2:

docker exec -it ros2_ws_container bash
ros2 run pubsub_pkg subscriber

Simulation

Gazebo:

docker exec -it ros2_ws_container bash
gazebo

RViz2:

docker exec -it ros2_ws_container bash
rviz2

TurtleBot3:

docker exec -it ros2_ws_container bash
export TURTLEBOT3_MODEL=burger
ros2 launch turtlebot3_gazebo turtlebot3_world.launch.py

Daily Use

Start:

docker compose up -d

Enter:

docker exec -it ros2_ws_container bash

Stop:

docker compose down

Project Structure

ros2-humble-nvidia-docker-template/
├── docker-compose.yml
├── Dockerfile
├── ros2_ws/
│   └── src/
│       └── pubsub_pkg/
└── README.md

The ros2_ws folder is mounted, so any file you edit on your host is live inside the container without rebuilding.

Author

Vignesh Suresh