ForzaETH Race Stack -- Scaled Autonomous Head-to-Head Racing on Fully Commercial off-the-Shelf Hardware

TL;DR

The ForzaETH Race Stack presents a complete, open-source autonomous racing stack designed for 1:10 scale F1TENTH hardware. It integrates state estimation with dual localization backbones, opponent detection and tracking, global and local planning, and a modular control framework to enable robust Time-Trials and Head-to-Head racing on commercial hardware. The system achieves competitive lap times, reliable opponent handling, and scalable performance across multiple tracks, while demonstrating low Latency and onboard computation. By providing reproducible hardware and software, it lowers barriers to entry for research groups and enables exploration of multi-opponent racing and future scale-up to full-size autonomous systems.

Abstract

Autonomous racing in robotics combines high-speed dynamics with the necessity for reliability and real-time decision-making. While such racing pushes software and hardware to their limits, many existing full-system solutions necessitate complex, custom hardware and software, and usually focus on Time-Trials rather than full unrestricted Head-to-Head racing, due to financial and safety constraints. This limits their reproducibility, making advancements and replication feasible mostly for well-resourced laboratories with comprehensive expertise in mechanical, electrical, and robotics fields. Researchers interested in the autonomy domain but with only partial experience in one of these fields, need to spend significant time with familiarization and integration. The ForzaETH Race Stack addresses this gap by providing an autonomous racing software platform designed for F1TENTH, a 1:10 scaled Head-to-Head autonomous racing competition, which simplifies replication by using commercial off-the-shelf hardware. This approach enhances the competitive aspect of autonomous racing and provides an accessible platform for research and development in the field. The ForzaETH Race Stack is designed with modularity and operational ease of use in mind, allowing customization and adaptability to various environmental conditions, such as track friction and layout. Capable of handling both Time-Trials and Head-to-Head racing, the stack has demonstrated its effectiveness, robustness, and adaptability in the field by winning the official F1TENTH international competition multiple times.

Figures (34)

• Figure 1: The physical ForzaETH autonomous racecar running the proposed ForzaETH Race Stack. On the right, an overtaking maneuver during the ICRA23 F1TENTH race.
• Figure 2: Comprehensive overview of the autonomous racing platform's hardware architecture. This figure presents an exploded view of the racecar, highlighting its key components as well as their integration.
• Figure 3: Architecture overview of the proposed ForzaETH Race Stack following the See-Think-Act paradigm. It highlights the interplay and interconnectivity of autonomy modules and hardware and illustrates how upstream tasks have knock-on effects that influence the subsequent autonomy modules. This depiction is inspired by betz_ar_survey, but emphasizes the importance of State-Estimation for autonomous racing, as a standalone autonomy module within Perception. Further, it is depicted how the ForzaETH Race Stack can switch seamlessly between the physical robot and the simulation environment.
• Figure 4: In \ref{['fig:frames']}, the frames of reference used for the ForzaETH Race Stack are shown. The inertial map frame is the reference frame for the global Cartesian coordinates used across this work. The other frames are rigidly attached to the car, with the body frame base_link at the center of the rear axle, the two sensor frames laser and imu at the center of the respective sensors. A representation of the two used coordinate systems is further shown in \ref{['subfig:cart_coord']} and \ref{['subfig:frenet_coord']}, with a reference trajectory in blue. In \ref{['subfig:cart_coord']}, the origin is represented by the red and green arrows, as it corresponds to the inertial map frame. The body frame base_link is further represented with a pair of red-green arrows, centered in the position of the car. The car's velocity is further represented in both the body frame, in \ref{['subfig:cart_coord']}, and in Frenet coordinates in \ref{['subfig:frenet_coord']}. \ref{['subfig:frenet_coord']} further shows the Frenet coordinate system's axis of the Cartesian system, with the origin and a few example points on the reference axes shown in black to facilitate the reader.
• Figure 5: An overview of the proposed state estimation system architecture. The state estimation module incorporates velocity estimation and localization with respect to the pre-mapped racetrack and aggregates this information in a final car state odometry output both in Cartesian and in Frenet coordinates.