Skip to content

Setup of the "ubm-f1tenth" environment

Linux/Mac

Install Foxglove studio here and create an account.

To remove the corrupted configs and restore the docker CLI config to default Run the following command: 

sudo rm ~/.docker/config.json
Now clone the repository and enter its folder. To obtain the docker image you can pull it from our dockerhub account, first login with: 
docker login -u ubmdriverless -p <redacted>
then to pull: 
docker pull ubmdriverless/f1tenth:latest
Alternatively you can build the image yourself with: 
docker build -t f1tenth-ubm-sim -f Dockerfile .
This will take a long time, around 10-30 minutes. You only have to do it once! Now start the docker container. 
docker compose up
And in as many other terminals as you want to access the simulation run: 
docker exec -it ubm-f1tenth-sim-1 /bin/bash
In one of these terminals bring up the simulation: 
launch sim.py
This command will launch the simulator and open a websocket at ws://localhost:9090

Now that the simulator is up, you can see it by connecting to the websocket at localhost:9090. Open Foxglove studio, log in and select Open connection... on the left panel. Then select Rosbridge in the popup window and make sure that the address is ws://localhost:9090, then click Open.

The main window is the 3D Panel, here you will see a 3D rendering of the data being published by ROS. To view the car click on the 3D panel, then on Panel in the left side-bar, here you can select which topics to view, make sure that Display frame is set to map, then scroll down to Custom layers and click on the three dots to add a URDF (this will be the 3D model of the car), in the URDF settings select Topic as source and /ego_robot_description in the Topic value.

If you are seeing a bunch text and arrows on top of the car, those are the car's transforms, you can hide them in the Transforms section.

Setting the Color by value of the /scan topic to range is advised.

For more info on how Foxglove works refer to the official documentation.

You will see the car and the track!
In another one of the Visual Studio Code terminals start up the control stack: 

launch control.py
In another one of these terminals send virtual telecommand signals: 
press_circle
This starts the pure pursuit algorithm and the car will move. The telecommand to stop the car is: 
press_L1
To shut it down, go to the first terminal, the one you started the container with "docker compose up" and press ^C. Next time you wish to code, open VSCode in the correct directory, start docker and connect with Tiger VNC.

Windows

For compatibility, it runs within WSL. This is a linux virtual machine provided by Microsoft that runs with near bare metal performance. Open a powershell prompt and install WSL: 

wsl --install
To make sure wsl uses Ubuntu as the default distribution: 
wsl -s Ubuntu
Now open wsl by typing: 
wsl
By default, you are put in a mount on your C: drive. For read/write performance, we will create a folder in your linux home. 
cd ~
mkdir  F1Tenth
cd F1Tenth
Now we set up github to clone the repository in this directory using SSH keys, since passwords are deprecated. 
ssh-keygen -t ed25519 -C "your_email@example.com"
Press enter to accept the default file location. Set a password.
Copy your SSH key: 
cat ~/.ssh/id_ed25519.pub
Go to https://github.com/settings/ssh/new and name your key, then paste it and click "Add SSH key". Now you can clone the repo: 
git@github.com:ubm-driverless/ubm-f1tenth.git
Install Foxglove here and create an account.

Do not close the powershell but go back to your Windows desktop, we have Docker Desktop and Visual Studio Code to install: https://docs.docker.com/desktop/install/windows-install/
https://code.visualstudio.com/download
Open Docker Desktop, go to general settings and ensure that "Use the WSL 2 based engine" is checked. Go back to your powershell. Navigate inside the repo and open it in Visual Studio code. 

cd ubm-f1tenth/
code .
This is your development enviroment. Cool! Now to install ROS2 in a docker container, use a new terminal within Visual Studio Code. This ensures it is the correct directory and also in WSL. It's perfectly fine to run this in a new terminal if the vs code one is not working. It might be solved by rebooting.

To obtain the docker image you can pull it from our dockerhub account, first login with: 

docker login -u ubmdriverless -p <redacted>
then to pull: 
docker pull ubmdriverless/f1tenth:latest
Alternatively you can build the image yourself with: 
docker build -t f1tenth-ubm-sim -f Dockerfile .
This will take a long time, around 10-30 minutes. You only have to do it once! Now start the docker container.

Now start the docker container. 

docker compose up
And in as many other terminals as you want to access the simulation: 
docker exec -it ubm-f1tenth-sim-1 /bin/bash
In one of these terminals bring up the simulation: 
launch sim.py
This command will launch the simulator and open a websocket at ws://localhost:9090

Now that the simulator is up, you can see it by connecting to the websocket at localhost:9090. Open Foxglove studio, log in and select Open connection... on the left panel. Then select Rosbridge in the popup window and make sure that the address is ws://localhost:9090, then click Open.

The main window is the 3D Panel, here you will see a 3D rendering of the data being published by ROS. To view the car click on the 3D panel, then on Panel in the left side-bar, here you can select which topics to view, make sure that Display frame is set to map, then scroll down to Custom layers and click on the three dots to add a URDF (this will be the 3D model of the car), in the URDF settings select Topic as source and /ego_robot_description in the Topic value.

If you are seeing a bunch text and arrows on top of the car, those are the car's transforms, you can hide them in the Transforms section.

Setting the Color by value of the /scan topic to range is advised.

For more info on how Foxglove works refer to the official documentation.

You will see the car and the track!
In another one of the Visual Studio Code terminals start up the control stack: 

launch control.py
In another one of these terminals send virtual telecommand signals: 
press_circle
This starts the pure pursuit algorithm and the car will move. The telecommand to stop the car is: 
press_L1
To shut it down, go to the first terminal, the one you started the container with "docker compose up" and press ^C. Next time you wish to code, start WSL, open VSCode in the correct directory, start docker and connect with Tiger VNC.

Useful Information

When to Use colcon build

  • After Cloning a New ROS2 Package or Project: If you’ve cloned a new ROS2 package or project into your workspace, run colcon build to compile it.
  • After Modifying Code or Configuration Files: If you make changes to source code or configuration files, you need to recompile using colcon build.
  • After Modifying CMakeLists.txt or package.xml: Any changes to these files require recompilation.
  • When Adding New Dependencies: If you add new dependencies, use colcon build to update them.
  • To Perform a Clean Build: Use colcon build --cmake-clean-cache to remove the CMake cache and rebuild from scratch.

Command:

colcon build

Useful Options:

  • --packages-select <package_name>: to compile only a specific package.

Additional Notes

  • Rebuild Docker Environment: Rebuild the environment only if modifications are made to the Dockerfile.

  • Change Map: To change the map used in the simulation, modify the sim.yaml file located in the f1_tenth_gym_ros directory.

  • Control and Speed Profiles:

  • Modify control.py only in the TUM_node.
  • Create a speed profile (only if it doesn't exist already).
  • Adjust lines 21 and 22 in raceline.py if needed.

When changing a map change: control.py, raceline.py, f1_tenth_gym_ros/config/sim_yaml, launch_config.yaml