Abstract
This is a continuation of a successful Safety21 project on developing a training community for engineering and ethical skills for developing future autonomous vehicles. This project includes three components - (1) autonomous driving course development with a 1/10th-scale autonomous racecar where students learn advanced algorithms and software development for perception, planning and control of autonomous driving; (2) Community Activities spanning 80 universities which have one or more F1tenth platforms and participate in the international autonomous racing competitions. We will host a minimum of 5 competitions in the top robotics, transportation and cyber-physical systems conferences; (3) Development of an ethical framework for using machine learning in life-critical systems.
Contribution: An autonomous vehicle hardware platform, called F1TENTH, is developed for teaching autonomous systems hands-on. This project will design and develop the education modules and software stack for teaching at various educational levels with the theme of ``racing" and competitions that replace exams.
Background: College-level robotics courses often focus on theory, while most hardware platforms for robotics teaching are low-level toys aimed at younger students at middle-school levels. The F1TENTH robotic race car fills the gap between research platforms and low-end toy cars and offers the hands-on experience in learning the topics in autonomous systems.
Intended Outcomes: The F1TENTH vehicles offer a modular hardware platform and its related software for teaching the fundamentals of autonomous driving algorithms. From basic reactive methods to advanced planning algorithms, the teaching modules enhance students' computational thinking through autonomous driving with the F1TENTH vehicle.
Application Design: Over 80 universities have adopted the teaching modules for their semester-long undergraduate and graduate courses for multiple years. Student feedback is used to analyze the effectiveness of the F1TENTH platform. This project's focus is to maintain and grow this community through education, outreach and K-12 training events.
Findings: More than 80% of the students strongly agree that the hardware platform and modules greatly motivate their learning, and more than 70% of the students strongly agree that the hardware enhanced their understanding of the subjects. The survey results show that more than 80% of the students strongly agree that the competitions motivate them for the course.
Description
Timeline
Strategic Description / RD&T
A new course for teaching hands-on autonomous systems with modular autonomous vehicle hardware and software has been created. The hypothesis is that autonomous driving fundamentals must be taught in combination with actual hardware to prepare the students for industry and academia jobs. This combination will enhance the students' computational thinking regarding the software and their systems thinking regarding the whole autonomous vehicle. This is because the students are allowed for repeated testing and iteration and have the affordance of a physical device to learn as opposed to on-screen simulation only.
Furthermore, it is hypothesized that by teaching autonomous driving in a competitive environment called Autonomous Racing, the motivation and fascination for learning in the field of autonomous vehicles and programming can be kept higher. The idea behind this variation of competition-based learning is to have three races in the course that incentivizes the students while not using rankings for grading, with the goal to teach more than in comparison to a standard class. Here is the course contents - https://tinyurl.com/F1TENTH-22Schedule
Deployment Plan
We developing https://courses.f1tenth.org/ for online offerings of this curriculum.
The course will be taught again in January-May 2024. More details on the overall project is at https://f1tenth.org
We have hosted 13 international autonomous racing competitions - https://f1tenth.org/race
The 12th competition was attended by over 150 participants from 23 different teams. All teams built the same reference platform and competed based on their autonomous racing algorithms. The next races will by at
2023-24 F1Tenth Races - https://f1tenth.org/race
1. CPSweek’23, San Antonio - RACE 11 (May’23)
2. ICRA’23, London - RACE 12 (June’23)
3. IEEE IV’23, Alaska - RACE 13 (June’23)
4. IROS’23, Detroit - RACE 14 (Oct’23)
5. ICCAS’23, 2nd Korea Race- RACE 15 (Oct’23)
6. ICRA’24, Yokohama, Japan- RACE 16 (May’24)
7. IEEE IV’24, Korea - RACE 17 (June’24)
8. IROS’24, UAE - RACE 18 (Oct’24)
9. IEEE ITSC'24, Canada. - RACE 19 (Oct'24)
We are also organizing a conference track at the IEEE Serious Open Source summit https://events.bizzabo.com/549239
Expected Outcomes/Impacts
The course instructors will develop 30 reference platform vehicles and demonstrate them working at high speeds of 8-15mph in corridors.
We will host 5 international competitions with over 80 academic partners worldwide.
Expected Outputs
Theme I: Safe Autonomy - This thrust will enable AV controllers that combine the performance and generalization abilities of machine learning with the safety guarantees afforded by formal and semi-formal verification. Researchers in this theme develop fast verification methods that scale to run in real-time on-board the vehicle through a combination of formal methods and testing. A cloud-based simulator will enable scalable verification which combines robust testing and falsification with reachability analysis for real systems.
Theme II: Efficient Autonomy - Researchers in this theme develop the hardware and software architectures for power-efficient and timing-guaranteed execution of autonomy algorithms. These include computer vision, motion planning, and neural network inference engines.
Theme III: Coordinated Autonomy - This thrust will enable a fleet of AVs to coordinate on-the-fly to achieve fleet-wide safety, higher transportation network efficiencies and enable exploration of new mobility and ridesharing services.
Theme IV: Secure Autonomy - Researchers in this thrust develop models of cyber-physical attacks, and resilient estimation and control schemes to guard against them and mitigate their effects in the field.
TRID
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.
Individuals Involved
| Email |
Name |
Affiliation |
Role |
Position |
| rahulm@seas.upenn.edu |
Mangharam, Rahul |
University of Pennsylvania |
PI |
Faculty - Tenured |
Budget
Amount of UTC Funds Awarded
$
Total Project Budget (from all funding sources)
$100000.00
Documents
Match Sources
No match sources!
Partners
| Name |
Type |
| Carnegie Mellon University |
Deployment & Equity Partner Deployment & Equity Partner |
| The Autoware Foundation |
Deployment & Equity Partner Deployment & Equity Partner |
