Rove is a robot developed by the Capra team at ÉTS. Utilizing ROS2 Humble, Rove is designed for advanced applications in search and rescue robotics.
Working in a dev container will allow you to have the same environment as the CI and make sure that your code will work on another computer. It will also allow you to easily switch package version and test things without breaking your computer.
- Install Docker Desktop
- Install Visual Studio Code
- Install Xserver (We recommend VcXsrv)
- Install the Remote - Containers extension in VSCode
- Clone and open the repository
- Click on the green button in the bottom left corner of VSCode and select "Remote-Containers: Reopen in Container" or use the command palette to do it.
- Wait for the container to build
- Start Xserver with the
-nowgl
option (double click on the shortcut to open it if you use VcXsrv)
Same as the windows installation, step 3 and 8 can be skipped.
Replace the DISPLAY environment variable in the .env file
echo DISPLAY=$DISPLAY
To be able to use the controller node, the user need read/write permissions on the inputs. Example:
cat /dev/input/event0
Suggestion: Configure docker to be able to run as non-root user https://docs.docker.com/engine/install/linux-postinstall/#manage-docker-as-a-non-root-user
To install ROS2 and vcs natively: https://docs.ros.org/en/humble/Installation/Ubuntu-Install-Debians.html
Note: At the step sudo apt install ros-humble-desktop
do sudo apt install ros-humble-desktop-full
instead.
Before installing the robot software, you need to install the simulator : https://gazebosim.org/docs/harmonic/install_ubuntu/#binary-installation-on-ubuntu
git clone https://github.com/clubcapra/rove.git
cd rove
vcs import src < rove.repos
echo "export GZ_VERSION=harmonic" >> ~/.bashrc && source ~/.bashrc
sudo rosdep init
rosdep update
rosdep install --from-paths src --ignore-src -r -y
colcon build --symlink-install
source install/setup.bash
If you intend to control the robot you might need to add the hardware package to the build (If you aren't sure, you probably don't need it):
vcs import src < rove_hw.repos
rosdep install --from-paths src --ignore-src -r -y
colcon build --symlink-install
source install/setup.bash
IF YOU ARE RUNNING IN WSL: do this command
export LIBGL_ALWAYS_INDIRECT=0 export LIBGL_ALWAYS_SOFTWARE=1
Do this command to run the gazebo simulation with physics enabled :
ros2 launch rove_bringup sim.launch.py
source install/setup.bash
ros2 launch rove_bringup rove_controller_usb.launch.py
source install/setup.bash
ros2 launch rove_bringup rove_controller_bluetooth.launch.py
ros2 launch rove_bringup vectornav.launch.py
Launch the gripper controller:
ros2 launch robotiq_description robotiq_control.launch.py
Close the gripper (set to 1m to reduce command lenght):
ros2 action send_goal /robotiq_gripper_controller/gripper_cmd control_msgs/action/GripperCommand "{command:{position: 1, max_effort: 1.0}}"
Open the gripper:
ros2 action send_goal /robotiq_gripper_controller/gripper_cmd control_msgs/action/GripperCommand "{command:{position: 0, max_effort: 1.0}}"
View the gripper in RViz:
ros2 launch robotiq_description view_gripper.launch.py
You can find the json configuration file in utils/ui/capra_ui.json. You need to load it into foxglove. On the device that you want to be connected (a.k.a. the jetson), you need to run the following command :
You can make it as a service and start at boot :
./utils/install.sh
Otherwise, to start use the user interface with a development laptop, you can run it by only launching the launchfile :
ros2 launch rove_launch_handler launch_handler.py
To add a package for Rove, create it using the ROS2 command (Creating Your First ROS2 Package). Name it starting with rove_
to ensure Git tracking. For non-Rove specific packages, create a separate repository and add it to rove.repos
.
Update Rove-specific packages directly in this repository. For Capra-related or external packages, update their references in rove.repos
. Change the Git branch in rove.repos
as needed and apply updates with vcs import src < rove.repos
.
Bringup: This essential package is responsible for initializing the rover in real-world settings. It activates all necessary subsidiary packages tailored for specific operational scenarios.
Description: This package is specifically designed for detailing the rover's structure in URDF (Unified Robot Description Format).
Gazebo: This package facilitates the simulation of the rover within the Gazebo environment, providing a virtual testing ground.
Hardware: Dedicated to initializing all sensors and actuators on the rover, this package is pivotal for operational readiness. It is specifically designed for use with the actual hardware and is not suitable for development machines.
Navigation: This is a wrapper package that integrates navigation functionalities, primarily based on the nav2 framework.
NLP (Natural Language Processing): This package manages the deployment of the NLP server and chatbot, facilitating advanced communication and processing capabilities.
SLAM (Simultaneous Localization and Mapping): This package serves as a comprehensive wrapper for SLAM operations, incorporating tools like slam_toolbox, rtab map, and sensor filters for effective environment mapping and rover localization.
It's possible to run the entire project into multiple docker containers. Each container can be run independently and are built using the following structure :
To update the UML diagram, create a new encoded link on the PlantUML website. Copy the existing UML, make your changes, and then update the Markdown file with the new link.