Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
126 tokens/sec
GPT-4o
47 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Autonomous Vehicles on the Edge: A Survey on Autonomous Vehicle Racing (2202.07008v1)

Published 14 Feb 2022 in cs.RO and cs.SE

Abstract: The rising popularity of self-driving cars has led to the emergence of a new research field in the recent years: Autonomous racing. Researchers are developing software and hardware for high performance race vehicles which aim to operate autonomously on the edge of the vehicles limits: High speeds, high accelerations, low reaction times, highly uncertain, dynamic and adversarial environments. This paper represents the first holistic survey that covers the research in the field of autonomous racing. We focus on the field of autonomous racecars only and display the algorithms, methods and approaches that are used in the fields of perception, planning and control as well as end-to-end learning. Further, with an increasing number of autonomous racing competitions, researchers now have access to a range of high performance platforms to test and evaluate their autonomy algorithms. This survey presents a comprehensive overview of the current autonomous racing platforms emphasizing both the software-hardware co-evolution to the current stage. Finally, based on additional discussion with leading researchers in the field we conclude with a summary of open research challenges that will guide future researchers in this field.

Citations (183)

Summary

  • The paper surveys the state of autonomous vehicle racing, detailing key software components (perception, planning, control), hardware platforms, and relevant competitions.
  • It highlights challenges and advances in high-speed perception systems, advanced planning strategies, control systems for dynamic racing conditions, and end-to-end learning approaches.
  • Autonomous racing is a key testbed for broader AV technology, highlighting future challenges like improving high-speed perception and optimizing the safety-performance balance.

Overview of Autonomous Vehicle Racing: A Survey of Current Developments

The paper "Autonomous Vehicles on the Edge: A Survey on Autonomous Vehicle Racing" presents a comprehensive exploration of the burgeoning field of autonomous racing, a specialized sector of the broader autonomous vehicles research domain. Authored by Johannes Betz et al., the paper provides an in-depth survey of the state of the art in autonomous racing technology, focusing on both software and hardware development. The discussions extend across various aspects such as perception, planning, and control, each critical for the successful implementation of autonomous functionalities in high-speed racing contexts. The survey delineates strong results and claims across these domains, highlighting key achievements and identifying persisting challenges within the research community.

Software Development in Autonomous Racing

The discussion of autonomous racing software is categorized into several components, namely, perception, planning, control, and end-to-end learning systems.

  • Perception Systems: The paper discusses challenges such as high-speed object detection and localization without traditional landmarks. The survey highlights adaptive techniques like SLAM, integration of DNNs for semantic segmentation, and predictive models that contribute to mapping and localization in dynamic conditions. These studies underpin algorithms that can function amidst high velocities and rapidly changing environments, ensuring robust vehicle performance.
  • Planning Strategies: In the planning domain, efforts are split between global path optimization for lap minimization and localized trajectory adjustments for obstacle avoidance. Recent advances in NMPC and game theory demonstrate how cutting-edge algorithms optimize racing performance by simulating adversarial conditions.
  • Control Systems: The survey comprehensively addresses path tracking under extreme handling conditions using classical and model predictive control (MPC) strategies. Here, algorithms are developed to adapt to high dynamic vehicle states, reinforcing timely and precise control over vehicle trajectories.
  • End-to-End Learning: The paper elaborates on partially integrated versus full end-to-end systems facilitated by DNN and RL, evaluating their application in seamlessly converting perception to control actions. These insights illustrate how end-to-end frameworks could potentially redefine adaptive race strategies.

Autonomous Racing Hardware and Competitions

A significant portion of the paper describes various autonomous racing platforms, from small-scale testbeds to full-sized race cars. These platforms offer diverse test conditions that help researchers validate theoretical models through competitions like F1TENTH and Roborace, which are pivotal in refining vehicle autonomy algorithms under quasi-real-world conditions.

Implications and Future Directions

The survey underscores the importance of autonomous racing as a testing ground for broader autonomous vehicle applications. The high-speed, high-stakes environment provides unique insights to enhance safety, improve algorithm efficiency, and push the boundaries of current vehicular capabilities. For the theoretical domain, this might involve developing advanced AI models capable of learning from vast datasets and refining predictive models better suited for human-autonomous interaction.

The paper concludes by outlining several open research challenges, suggesting that the future of autonomous racing will likely involve improving high-speed perception, advancing multi-agent planning, addressing real-time computing demands, and refining the balancing act between safety and performance. By presenting these open areas, the survey invites further exploration and development, encouraging ongoing discourse in both academic and applied research circles.

This extensive survey is not only a repository of achieved milestones but also a strategic guidepost for researchers aiming to contribute to the future of autonomous vehicle systems. As we progress, the findings and methodologies from autonomous racing continue to offer invaluable lessons for the broader spectrum of intelligent vehicle development.