On 7 July 2022, the hallowed halls of TU Wiens’ famous Kuppelsaal were put to an unusual use. Our computer engineering students and their instructors turned them into a steep race track for their self-built autonomous racing cars.
Radu Grosu, Head of the Research Unit for Cyber-Physical Systems initiated the course in 2020 inspired by the F1TENTH Autonomous Racing which has been developed by Rahul Mangharam at the Safe Autonomous Systems Lab, University of Pennsylvania and was published under CC-NC-SA 4.0 license.
For instructors Andreas Brandstätter and Radu Grosu this year’s course was exceptional. Due to the pandemic, final races couldn’t be held in the previous years – it was the first race where all students were able to put their cars to the test. In the annual course, students learn to master all hardware and software skills required to build a fully autonomous racing vehicle. Current technologies in self-driving cars and mobile autonomous systems are being taught hands-on: During many weeks, students not only build their vehicles from scratch but develop and implement racing strategies, computer vision, and machine learning algorithms that give their team an edge in the race that concludes the course.
The teams ‘GIMP’, ‘tyre my beauty’, and ‘Frankencar’ gave their all at the racing track. To further develop the course format and use international synergies, Nicolas Baumann from ETH Zürich and Jaroslav Klapálek from CTU joined the final race with colleagues and students. Special thanks also goes to the highly motivated tutors of the course, who spared no time and effort to support the racing teams.
The purchase of the F1TENTH cars was made possible by the BMBWF CPS/IoT-Ecosystem Infrastructure Award. Racing is contagious – the course’s best students usually take it to the next step at the international F1TENTH Grand Prix. TU Wien Informatics’ team made 3rd place in 2022 and won in 2021.
