TurtleBot4 Simulator (ROS 2 Jazzy + Gazebo Harmonic) — Docker

A reproducible environment for simulating TurtleBot4 using Gazebo (Harmonic). This setup supports SLAM (slam_toolbox) for map generation, saving maps, and running Localization + Navigation2 (Nav2).

Note: Tested on Ubuntu 24.04 with NVIDIA GPU (OpenGL 4.6) using --gpus all.

Project Structure

turtlebot4_jazzy/
├── sim/                    # Dockerfile, run scripts and Gazebo world
├── maps/                   # Saved maps (yaml/pgm)
├── params/                 # Nav2 and waypoint parameters
├── scripts/                # Mission, diagnostics and benchmark helpers
├── factory_integration/    # OPC-UA/factory UI experiments
├── vision/                 # Vision datasets/assets
└── README.md               # Documentation

Requirements

• Docker and NVIDIA Container Toolkit (for GPU acceleration)
• X11 Server active (required for RViz2 and Gazebo GUI)
• Internet connection (to download worlds and ROS/Gazebo packages)

Build

Build the Docker image:

cd sim
docker build --no-cache -t turtlebot4:jazzy .

Run Container

Start the container with GPU and X11 forwarding enabled:

./sim/run_docker.sh

Inside the container, source the ROS environment:

source /opt/ros/jazzy/setup.bash

Usage

1. SLAM + Nav2 + RViz (Mapping)

To start the simulation with SLAM and mapping enabled:

ros2 launch turtlebot4_gz_bringup turtlebot4_gz.launch.py \
  slam:=true nav2:=true rviz:=true

In RViz:

• Set Fixed Frame = map.
• Add LaserScan (/scan) and Map (/map) displays if they are not already present.

Verification:

ros2 topic hz /scan
ros2 topic echo /map --once --qos-durability transient_local

You can now drive the robot using the Teleop panel in the Gazebo GUI or via ros2 run teleop_twist_keyboard teleop_twist_keyboard.

2. Saving the Map

If the container was started with the volume mount -v $PWD/maps:/root/maps (default in run_docker.sh), you can save the generate map to the host machine:

ros2 run nav2_map_server map_saver_cli -f /root/maps/warehouse

This will generate maps/warehouse.yaml and maps/warehouse.pgm.

3. Localization + Nav2 (Navigation)

To run navigation with a pre-existing map (disable SLAM):

# Optional: Launch simulation separately if needed
# ros2 launch turtlebot4_gz_bringup turtlebot4_gz.launch.py

# Launch Localization (AMCL) providing the map path
ros2 launch turtlebot4_navigation localization.launch.py map:=/root/maps/warehouse.yaml

# Launch Nav2
ros2 launch turtlebot4_navigation nav2.launch.py

In RViz:

1. Use 2D Pose Estimate to set the initial robot position.
2. Set a destination using 2D Goal Pose or the Navigation2 panel.

Troubleshooting

• Known upstream issues and B002 notes: See KNOWN_ISSUES.md for the TurtleBot4/Nav2/network issues that influenced this project.

• Broken TF (map -> odom -> base_link): Run view_frames to verify the transform tree is connected.

  ros2 run tf2_tools view_frames

• Map not appearing: Ensure the QoS profile is set to TRANSIENT_LOCAL.

  ros2 topic echo /map --once --qos-durability transient_local

• GUI Freezing / Crashing: Verify GPU acceleration inside the container:

  apt-get update && apt-get install -y mesa-utils
  glxinfo | grep -E "OpenGL renderer|OpenGL version"