Base for robotics course

Prerequisites

• Docker / Docker Desktop
• Docker compose (included in the docker desktop?)
• If on ubuntu with Nvidia GPU: Nvidia-docker2

Docker

Building the image

Run following command at docker folder of this repo to build the docker image.

docker build -t turtlebot:latest -f Dockerfile ../

Breakdown of the build command

docker build

is the straight forward the command for building.

-t <NAME>:<TAG>

Argument -t defines the name and a tag of the image which are separated by colon. If no tag is specified, it will be set to latest by default. Tags are an easy way to separate different versions of the images.

-f Dockerfile

Argument -f defines the Dockerfile to be used. Works fine with relative paths.

../

The final piece of the command tells the context where docker will start looking for files. In our case for example in the Dockerfile we do COPY ros2_ws/src ... If the build command was ran from the docker folder with the ../ at the end it means that the COPY command will copy data from robotics_course folder.

VPN may cause issues. --network=host fixes the issue but is not the safest way to handle it.

Running the container(s)

There are two ways to run your image. Straight from command line or with the help of docker-compose.

Ubuntu: To enable usage of the GUI

xhost +local:`docker inspect --format='{{ .Config.Hostname }}' turtlebot`

---

Running on windows

• WSL2 must be installed (should be by default on Windows 11).

• Install XLaunch to allow usage of Gui.

• Run ipconfig on your host windows machine and set your local network IP address to the docker-compose.yml replacing the <IP> part with the one you find there. ipconfig will give you more options. Try out the values corresponding to the IP V4 Address fields under the following sections: Wireless LAN Adapter Wi-Fi, Ethernet adapter vEthernet (Default Switch), Ethernet adapter vEthernet (WSL).

• Start XLaunch (VcXsrv Windows X Server).

• Press next until you get to the Extra settings tab.

• Deselect the option Native OpenGL.

• Select Disable access control.

---

Simplest possible run command from command line:

docker run -it turtlebot:latest bash

Where docker run tells docker that we want to run an container

-it

Defines an interactive mode, which allows us to use the command line properly inside the image.

turtlebot:latest

Defines the image which we want to run

bash

As last we define the command we want to run when the container is started. In this case a bash terminal is started. Here we could basically run any program we have in the image.

Running images from command line like this can be done but if you run more complex systems with docker you would have to add many arguments into the run command and that becomes quite quickly a mess.

Docker compose

To solve the issue of messy run command we can use docker compose. For compose we have to create a docker-compose.yml file which will include all the containers we want to launch. This repo includes an example compose file which can be used as basis for your work on this course.

To start compose you have to navigate to the docker folder where the file is located at and then run the following command (depending on your version of docker). Rest of the commands on this document will be using the newer command structure so if you have the older version, you will have to add the dash into the commands, or update your docker.

docker-compose up

Or

docker compose up

This will start the container with the command specified in the docker-compose.yml file. You will not have access to the command line in that terminal unless you specify detach argument -d.

To shutdown the containers you can run

docker compose down

Also simple ctrl+C will do the job if you did not set the detach flag.

docker compose logs --follow

Accessing the command line of the container

To access the container via command line you can use

docker exec -it turtlebot bash

Where docker exec defines again the command for docker

-it

Is again making the session interactive and allows printing into the console.

turtlebot

Next we have to define the name of the container we want to get into

bash

And as last thing we define what command we want to execute in that container. In case that we want to access the command line we will use bash.

Other handy docker commands

To list all docker images on the device use

docker images

To list all the running containers

docker ps

To stop a docker container

docker rm <container name/ID>

To start stopped container

docker start <container name/ID>

To remove docker container (Cannot be anymore started by docker start command)

docker rm <container name/ID>

Sometimes you might run into situation where docker complains that the container cannot be started because other one with same name is running. To fix that you have to run docker stop and docker rm for the container.

Turtlebot simulation launch

1. Launch the turtlebot_sim with Gazebo Ignite and the basic maze world:

ros2 launch turtlebot4_ignition_bringup turtlebot4_ignition.launch.py world:=maze

2. Launch the turtlebot_sim with Gazebo Ignite, in mapping mode:

ros2 launch turtlebot4_ignition_bringup turtlebot4_ignition.launch.py slam:=true nav:=true world:=maze

3. Launch the turtlebot_sim with Gazebo Ignite, in localization mode, with Rviz:

ros2 launch turtlebot4_ignition_bringup turtlebot4_ignition.launch.py nav2:=true world:=maze map:=/maze.yaml localization:=true rviz:=true

Resources:

• https://github.com/turtlebot/turtlebot4/tree/humble
• https://github.com/turtlebot/turtlebot4_simulator
• https://turtlebot.github.io/turtlebot4-user-manual/overview/features.html