This repository provides a simplified software system designed to help new Formula Student Driverless teams get started more easily. To this end, selected components of the autonomous system developed by Mainfranken Racing are made publicly available here.
⚠️ Prerequisites This system requires ROS Noetic on Ubuntu 20.04.
Installing the dependencies:
./check_dependencies.shBuilding the workspace:
catkin buildThe repository contains all modules required to complete all disciplines. For each discipline, a dedicated launch file can be executed.
roslaunch driverless_daemon acceleration.launch # Discipline: Acceleration
roslaunch driverless_daemon autocross.launch # Discipline: Autocross
roslaunch driverless_daemon skidpad.launch # Discipline: SkidPad
roslaunch driverless_daemon trackdrive.launch # Discipline: TrackdriveAt minimum, the system requires an IMU, wheel speed sensors on all wheels, and either a stereo camera or a LiDAR sensor.
The configuration files for the individual modules are located within their respective packages. Depending on the setup, it is also possible to choose between LiDAR, stereo camera, or sensor fusion configurations.
ℹ️ Info The interfaces to the corresponding sensors or to a simulator (e.g. https://github.com/mfr-driverless/fssim-2021/tree/master) are not included in the repository and must be set up beforehand.
To help you get started, we also provide a publicly available dataset. The dataset is intended primarily to support state estimation development and to serve as a benchmark for evaluating and comparing different approaches.
The dataset is available here: Zenodo Dataset 🔗
It can largely be used with the software system provided here without requiring significant modifications.
Video demonstration ...