Team Scuderia Segfault from TU Wien delivers incredible performance and wins the 22nd F1TENTH Autonomous Grand Prix, held from December 16th to 19th as part of the IEEE Conference on Decision and Control (CDC 2024) in Milano, Italy. This year, the competition against top research institutes and renowned universities was very close, but TU Wien’s team managed to outperform every other car and secure a win!
Preparation for the race
In the weeks before the event, our team thoroughly prepared for the race. Based on the expertise from past participation in international races, they had learned what it needs to perform best, even under challenging and stressful conditions. They not only improved the racing algorithm and the car hardware to stay up to date with technological advancements but also further improved and tuned the race-line generation methodology. Furthermore, they developed a structured plan for handling all tasks at the on-site race. Beginning from the first identification of friction parameters, the mapping procedure, up to the stressful head-to-head races, which require fast decisions, they developed standard procedures. These procedures were further supported by according checklists. In the weeks before the event, our team performed extensive drills to get familiar with these procedures.
Mapping and testing day
The preparation really paid off, as the team could work very efficiently on all the required tasks of the first mapping and testing day. Within the assigned time slot, the team was able to quickly record the required data and already generate a first race-line to be tested within a couple of hours. The comparison to other teams showed, that some of them struggled to get their car running properly within the limited time that was available for testing. Over the night, our team consulted online with the other team members to prepare further optimizations for the following day, improve the performance of the race-line, and adjust the algorithm perfectly for the given track conditions.
Best performance in time-trials
The first part of the race was the time-trials, where the autonomous cars needed to show speed and reliability: The fastest lap, as well as the number of laps without crashes, are both recorded and determine the ranking for the following races against other cars. Our team was testing multiple different configurations and could thereby significantly improve the performance, compared to the test runs on the previous day. In the first heat, they decided on a fast but also reliable set of parameters, which resulted in a lap time of 7.288 seconds with 41 consecutive laps. Other teams were very close, so the team decided to push the car even closer to its limits for the second heat, which improved the lap time to 7.126 seconds and 42 laps in total. This was overall the best performance in both metrics and resulted in first place for the time trials. The second-fastest team in the time-trials was HiPeRT UNIMORE Racing with 7.296 seconds, followed by Unibo Motorsport with 7.314 seconds in third place.
Head-to-Head Tournament
The final day was all about head-to-head races, where two autonomous cars ultimately race for the win. The first pairing of the double-elimination races for our team was against HUN-REN SZTAKI, which was considerably slower and resulted in a clear win for Scuderia Segfault. Next was against HiPeRT UNIMORE Racing – the second-fastest team in the time-trials. It was clear that this race needed careful adjustments of all parameters and consideration of every detail, in order to drive fast and reliable. The team managed to hold their nerves and also win this race. In the semi-final, it was a classic against ForzaETH. This confrontation is already well known in the scene of F1TENTH and always promises a thrilling competition. This time it seems, that the team from ETH had some hardware problems from a severe crash in the days before. However, they managed to get their car again ready to drive, but with degraded performance, which resulted in a clear win for our team, advancing to the finals. The finale against HiPeRT UNIMORE Racing was executed as a best-of-three race, resulting in an ultimate win for Scuderia Segfault.
About the team
Scuderia Segfault is TU Wien Informatics’ team for research on autonomous cars and autonomous racing in the F1TENTH series, consisting of Felix Resch, Luigi Berducci, Moritz Christamentl, Agnes Poks, Philipp Mandl, Elisa Di Cristo, Larisa Clement, Mihai-Teodor Stanusoiu, Monika Farsang, Nino Wegleitner, Philipp Gratzer, Piet Kaul, Samuel Lechner, under the supervision of PreDoc Researcher Andreas Brandstätter and Professor Radu Grosu from the Research Unit for Cyber-Physical Systems. A part of the team traveled to Milano to participate in the race; others contributed to preparation and development beforehand and online remote support during the race.
Thanks to our supporting partners
Our team is grateful for the generous support from Liquid AI and TTTech Auto. Some team members are supported by project NimbleAI. The team was further supported by HTU and Magna International.
