This is a guide to make the setup process easier for everyone, some may find the steps trivial but it's to get everybody on board before we start the development.
https://www.docker.com/products/docker-desktop/
Possible errors
- you have to enable hardware assisted virtualization on windows look here
- if you have an older version of operating system you may want to install an older version of docker look here
This Docker container is set up for running a robotic simulation environment with the following specifications:
ROS Version: The container is configured to run with ROS (Robot Operating System) in the "Humble" release.
Gazebo Version: Gazebo, a simulation tool, is included in version 11.10
Operating System: The base operating system is Ubuntu 22.04.3 LTS.
Download github desktop and clone this repository (arm). By doing open in finder/file explorer you will be able to locate the folder GitHub on your computer. Save the absolute path of this folder. (it's needed in the next steps)
From the githb desktop app go to current branch (which is going to be set on main) and click New Branch. Name it with your username. (during development each branch is going to represent a feature)
From github desktop you can open the whole arm repository on vscode. (from the top bar: repository -> open in visual studio code)
- important note: in docker before executing the following commands you have to set in preferences -> resources -> file sharing the path to the github folder
MacOS:
docker run -p 6080:80 --security-opt seccomp=unconfined --shm-size=512m -v /users/<username>/<path_to_github>:/github --name roverchallenge vossgit/ros-roverchallenge:stable
Windows:
docker run -p 6080:80 --security-opt seccomp=unconfined --shm-size=512m -v C:\<absolute_path_to_github>:/github --name roverchallenge vossgit/ros-roverchallenge:stable
At this point you should get as output a bunch of RUNNING state lines and you can proceed.
- note: replace for example the whole string like <absolute_path_to_github> -> /user/doc/github
- very important note: do not remove or replace :/github because it's the path for the folder inside of the virtual machine
After running the container, you can access the graphical user interface (GUI) by opening a web browser and navigating to http://localhost:6080. The container exposes the GUI on port 6080, allowing you to interact with the simulation environment.
To restart the container when closed(you should always have the docker app running in the background) in the computer terminal/shell write:docker start roverchallenge
(you should always have the docker app running in the background) in the computer terminal/shell write:
docker start roverchallenge
docker exec -it roverchallenge /bin/sh
The -v /users/<username>/<path_to_github>:/github part of the Docker run command establishes a volume mount. This allows you to share data between your host machine and the Docker container. Here's a breakdown of this volume mount:
/users/<username>/<path_to_github>: with the actual path on your host machine that corresponds to the github folder, that way it will be accessible from the Docker container.
:/github: This is the path inside the Docker container where the shared data will be available. In this case, it's mounted at /github.
the following commands are going to be executed from a terminal inside your host machine:
After having entered the container with:
docker exec -it roverchallenge /bin/sh
We have to set the command line to use bash by writing
bash
let's source ROS 2
. /opt/ros/humble/local_setup.bash
navigate into
cd github/arm/setup/rovertest
where we are going to build the cmake files and install the packages
colcon build
. install/setup.bash
Possible errors
if the colcon build does not instantly work try to
colcon build --cmake-clean-cache
To visualize rviz and gazebo go to
you should see a vnc viewer screen. Connect.
the following commands are going to be executed inside the ubuntu gui from the vnc viewer on localhost:6080
to source ROS 2
. /opt/ros/humble/local_setup.bash
Now open a terminal with the terminator app
now we can navigate to the system folder of ubuntu (which is simply /) with
cd ../../
navigate into
cd github/arm/setup/rovertest
then let's setup the packages with
. install/setup.bash
NEW! START GAZEBO+RVIZ+MOVEIT Launch in 3 different terminals in this order the following commands:
ros2 launch arm_description gazebo.launch.py
ros2 launch arm_controller controller.launch.py
ros2 launch arm_moveit moveit_launch.py
Start rviz
ros2 launch arm_description display.launch.py
ros2 launch rover_model display.launch.py
Start gazebo
ros2 launch arm_description gazebo.launch.py
gazebo -s libgazebo_ros_init.so -s libgazebo_ros_factory.so myworld.world
and in another terminal
ros2 launch rover_model gazebo.launch.py
if gazebo does not start correctly / does not load the model
It might not start and give "waiting for serviece /spawn_entity" logs and in the end it could retourn "process has died". To fix that start this before running gazebo.launch.py in a different terminal
gazebo -s libgazebo_ros_init.so -s libgazebo_ros_factory.so myworld.world
then by keeping this process going, open another terminal (dont forget the recurring instructions(look at the last paragraph of this readme))
ros2 launch rover_model gazebo.launch.py
if it still does not start restarting the whole docker container often helps
In a terminal on the HOST MACHINE, these commands need to be run JUST ONCE (there is no need to re run these even after you restart the whole container)
echo "source /opt/ros/humble/local_setup.bash" >> ~/.bashrc
echo "source /github/arm/setup/rovertest/install/setup.bash" >> ~/.bashrc
source ~/.bashrc
sudo printf "source /github/arm/setup/rovertest/install/setup.bash\n" >> /home/ubuntu/.bashrc
source /home/ubuntu/.bashrc
(This way there is no need to write these two commands all the time we open a terminal
. /opt/ros/humble/local_setup.bash
. install/setup.bash
)
commands if you are on vnc UBUNTU
cd ../../
cd github/arm/setup/rovertest
commands if you are on the HOST MACHINE
bash
cd github/arm/setup/rovertest
colcon
colcon build
generally used in HOST MACHINE where we have our code editor. It's made to build the packages, we use it each time we edit the files inside rover_model (same for any other ros package). It updates the files inside of rovertest/build/rover_model.
Therefore we always want to execute this command from
cd github/arm/setup/rovertest