Windrunner

Windrunner is a model car with Ackermann steering. This was an exercise on bringing CAD models into a physics simulator like Gazebo and further integrating ROS controllers to operate the vehicle. This was my first project for the ENPM662 (Introduction to Robot Modeling) course in the Fall 2022 semester.

The above image is run at 2 times the actual simulation speed for demonstration purpose.

Description of contents

cad/ consists of the Solidworks part and assembly files of the robot and tool plus Blender files of the same to export in DAE file format for the meshes.

dockerfile/ consists of the necessary files to build the Docker image of this project. The image can then be used to run the container as an executable for simulating in Gazebo. Building the Docker image locally is not recommended, an up-to-date image is available on Docker Hub.

image/ just consists of the GIF image used on the README page.

windrunner/ is the ROS package for the Windrunner model car.

Running the project

This project can only run on Linux at the moment in two ways:

In Docker container (Recommended)

This is the easiest, most convenient way to run the project demo as an executable. Docker (Linux) installation is required, to install refer the documentation.

Pull the Docker image of this project from Docker Hub.

docker pull armdocks/windrunner

Run the container for running the Gazebo simulation.

docker run -it -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix armdocks/windrunner

Tip: The -it argument can be omitted from the above command for a less verbose, clean terminal output.

Note: The simulation might lag when the Gazebo window opens maximized, shrink the window to improve frame rate.

On local machine

This requires Ubuntu Focal (20.04) with ROS Noetic full desktop installation. Additional ROS packages that might be required are joint state and effort controller packages and the PyYAML Python package.

These additional packages can be installed by:

Joint state controllers package

apt install ros-noetic-joint-state-controller

Effort controllers package

apt install ros-noetic-effort-controllers

PyYAML package

pip3 install pyyaml

Assuming you have the catkin workspace setup,

Clone the repository to src/ folder of the target catkin workpace. Assuming this would be ~/catkin_ws/src/,

git clone https://github.com/armgits/windrunner.git ~/catkin_ws/src/windrunner_project

Make the Python files start.py and move.py in the src/ folder of windrunner package executable.

sudo chmod +x ~/catkin_ws/src/windrunner_project/windrunner/src/windrunner_publisher.py
sudo chmod +x ~/catkin_ws/src/windrunner_project/windrunner/src/windrunner_teleop.py

Now you can build and source the packages.

After a successful build, run the simulation in Gazebo.

This can be done in a single line using the shell script windrunner_circle.sh or windrunner_teleop.sh from dockerfile/ folder.

Demo for moving in a circular path (As seen in GIF)

~/catkin_ws/src/windrunner_project/dockerfile/windrunner_circle.sh

Teleoperation using keyboard input

~/catkin_ws/src/windrunner_project/dockerfile/windrunner_teleop.sh

Or manually,

In first terminal initialize the ROS master.

roscore

Demo for moving in a circular path

In a new, second terminal spawn the robot in Gazebo.

roslaunch windrunner windrunner_empty_world.launch

In a new, third terminal run the node to perform the task.

rosrun windrunner windrunner_publisher.py

For teleoperation using keyboard input

In a new, second terminal spawn the robot in Gazebo.

roslaunch windrunner windrunner.launch

In a new, third terminal run the node to perform the task.

rosrun windrunner windrunner_teleop.py