ForzaETH by Autonomous Racing Zürich

???? Take a look at our latest video showcasing our advanced overtaking capabilities! Using our Predictive Spliner algorithm, we’re able to plan overtakes across an entire lap by accurately predicting and adapting to the opponent’s racing line. https://lnkd.in/eMtSMjyY

????? Thrilled to have showcased our autonomous racing innovations at the Swiss Robotics Day in Basel last week! Partnering with Center for Project-Based Learning D-ITET, we engaged with a dynamic community of robotics enthusiasts, engineers, and industry leaders, sharing insights into our research and capabilities in autonomous vehicle technology. Events like these are a fantastic reminder of the power of collaboration and curiosity to drive technological advancements. A big thank you to everyone who stopped by to discuss the future of robotics with us!

Last week, the ForzaETH race team participated in the F1Tenth Foundation 21st F1TENTH Autonomous Grand Prix at IROS. ????? At the race, we were faced with extremely grippy carpet floor, requiring teams to extract 100% of performance from both their hardware and software. This was in contrast to the robustness tuning the team had been hard at work with, after experiences with slippery racetracks at past races. Faced with this "out of distribution" environment, the team worked hard on-site, leaving no stone unturned as we sought to shave tenths of seconds from our lap times. Our efforts resulted in a 7.14s lap over the roughly 35m track. This is an average speed of ~5m/s (18kph), significant considering the curvature of the track! The car accelerated at 8 m/s2 and braked at 20 m/s2 – accelerations that mimic full-scale racing. This was sufficient for us to qualify in 2nd place. Unfortunately, in the head-to-head shootout, the team's high-risk, high-reward strategy to minimize laptime bit us as we crashed out of our match which would decide if we would advance to the final. We ended up with a 3rd place finish, behind our competitors Lukasz Sztyber and Scuderia Segfault. Despite this, we still can be proud of our Race Stack, especially in head-to-head mode, where our car demonstrated advanced abilities to estimate, predict, and plan to overtake opponents - made possible by our recent Predictive Spliner publication. We thank the competition organizers Rahul Mangharam, Fran?ois Pomerleau, Chinmay Samak, Tanmay Samak, and look forward to the next race. Race Blog Post: https://lnkd.in/eAMPKxac Predictive Spliner Paper: https://lnkd.in/eR5CGpEd Open Source Race Stack: https://lnkd.in/eWjswFXU Our IROS Race Team: Neil Reichlin, Tian Yi Lim, Benedict Hildisch, Nadine Imholz, Nicolas Baumann, Edoardo Ghignone, Cheng Hu