Turtlebot4 Docker

This repo provides multiple containers to work with Turtlebot4 Robots in DiscoveryServer Mode.

This pacakge provides and utilizes following containers

• Turtlebot4-Discovery
  Provides the discovery service for the simulation

• Turtlebot4-Simulation
  Provides the Gazebo based simulation environment

• Turtlebot4-Rviz
  Provides rviz tools to interact with both simulation and physical robots

• Turtlebot4-workspace
  Provides a workspace that can be used to test code on both simulation and physical robots

This packages contains two compose files that handles environmental variable configurations for Discovery Server on behalf of the user, so the user can focus on development without configuration worries

• compose.yaml for simulated robot
• compose.physical for physical robot (assumes Nav2 & Slamtoolbox is running natively on the robot)

Information

Read more information on following documents

• Container Information
• Troubleshooting

Install Docker

Follow the official instructions here

Clone the repo

On the terminal run the following command to clone the repo

git clone https://github.com/CollaborativeRoboticsLab/turtlebot4-docker.git

Start the workspace

Enter the folder

cd turtlebot4-docker

Pull the latest docker containers

docker compose pull

Required environmental variable need to be in a .env file. An example.env file is available. Rename that file to .env and update the values as required.

Workspace via code-server

When the stack is up, the workspace container serves a web VS Code at http://127.0.0.1:8080.

• Default password: student (set via PASSWORD in compose)
• Workspace folder: /workspace (mounted from ./turtlebot4-workspace/workspace)
• Settings/extensions persist to:
  • ./turtlebot4-workspace/code-server/config → /root/.config/code-server
  • ./turtlebot4-workspace/code-server/data → /root/.local/share/code-server

To change the password, edit the PASSWORD env in compose.yaml under the turtlebot4-workspace service.

Use the Simulation robot

On the compose.yaml file, Change the following line to select mapping or localization.

command: ["ros2", "launch", "turtlebot4_gz_bringup", "turtlebot4_gz.launch.py", "nav2:=true", "localization:=true", "slam:=false", "use_sim_time:=true"]

• slam:=true will allow the simulated robot to map the room. use gazebo joystick option to move the robot
• nav2:=true along with localization:=true to navigate the robot. Default is warehouse

Start the docker containers

docker compose up

For GUI containers (simulation, rviz), allow X11 access on the host if needed:

# Allow local-root X11 access for GUI apps in containers
xhost +local:root

Using the Physical robot and remote pc

Mapping

On the robot run following command to run the mapping process. Use a name like office to save the file

ros2 launch turtlebot4_navigation slam.launch.py

Move the robot using joystick to map the room.

Navigation and Localization

On the robot run following command to run the localization process with a created map

ros2 launch turtlebot4_navigation localization.launch.py map:=office.yaml

And on another terminal run nav2

ros2 launch turtlebot4_navigation nav2.launch.py

Then use the 2D pose Estimate tool to align the robot's laser scan with the map. Finally give a Navigation goal via Nav2 Goal tool

Visualization on remote pc

Export following commands on the remote pc terminal. Make sure the ROS_DOMAIN_ID and ROS_DISCOVERY_SERVER matches the robot's configuration.

export RMW_IMPLEMENTATION=rmw_fastrtps_cpp
export ROS_SUPER_CLIENT=True
export ROS_DOMAIN_ID=10
export ROS_DISCOVERY_SERVER="10.42.0.1:11811;"

And then run the following command to open rviz

ros2 launch turtlebot4_viz view_navigation.launch.py

Visualization on remote pc via docker container

Update the compose-physical.yaml file with correct configuration on following parameters

- ROS_DOMAIN_ID=10
- RMW_IMPLEMENTATION=rmw_fastrtps_cpp
- ROS_DISCOVERY_SERVER=10.42.0.1:11811
- ROS_SUPER_CLIENT=1

Allow permission for UI interfaces from docker containers

# Allow local-root X11 access for GUI apps in containers
xhost +local:root

Start the docker containers

docker compose -f compose-physical.yaml up

Doing custom work

Use code-server to edit files under /workspace (mounted from ./turtlebot4-workspace/workspace). We recommend creating a ROS 2 colcon workspace inside /workspace:

# In the code-server terminal (already sourced via ros entrypoint)
mkdir -p /workspace/turtle_ws/src
cd /workspace
colcon build --merge-install

# Use the overlay
source /workspace/install/setup.bash
ros2 pkg list

Place your packages in /workspace/turtle_ws/src, build with colcon, and run as usual. The code-server terminal is ROS-ready (environment is sourced by the container entrypoint).