Learn-to-Race

We seek to compete in a variety of autonomous driving competitions, both virtual and physical, by leveraging and improving upon existing reinforcement learning and computer vision approaches and then publishing in the public domain.
Learn-to-Race was founded in 2020, as a capstone project within the Masters in Computational Data Science (MCDS) program, in the School of Computer Science at Carnegie Mellon University.

Advisors

Jonathan Francis

Jonathan Francis

Technical Lead

Bosch Center for AI

Dr. Jonathan Francis is a Research Scientist at Bosch Center for Artificial Intelligence and an affiliated researcher in the Language Technologies Institute and Robotics Institute, in the School of Computer Science, at Carnegie Mellon University; he is a researcher and technical lead on the Learn-to-Race initiative. Jonathan maintains a research focus in domain knowledge-enhanced representation learning, applied to robotics and autonomous driving, where, in the context of Learn-to-Race, he focuses on safe reinforcement learning in the pursuit of safer and more generalisable autonomous systems. Jonathan received his PhD from the School of Computer Science at CMU in 2022.
Eric Nyberg

Eric Nyberg

Director

School of Computer Science, Carnegie Mellon University

Noted for his contributions to the fields of automatic text translation, information retrieval, and automatic question answering, Nyberg holds a Ph.D. from Carnegie Mellon University (1992) and a B.A. from Boston University (1983). He is a recipient of the Allen Newell Award for Research Excellence (for his contributions to the field of question answering and his work as an original developer on the Watson project) and the BU Computer Science Distinguished Alumna/Alumnus Award. Eric currently directs the Master of Computational Data Science (MCDS) program. He is also co-Founder and Chief Data Scientist at Cognistx and serves on the Scientific Advisory Board for Fairhair.ai.
Anirudh Koul

Anirudh Koul

Mentor

Pinterest

Anirudh Koul is the Head of Machine Learning and Labeling Sciences at Pinterest. He serves as a mentor on the Learn-to-Race affiliated capstone projects, in the Master's in Computational Data Science (MCDS) program, from the Language Technologies Institute, in the School of Computer Science, at Carnegie Mellon University.
Siddha Ganju

Siddha Ganju

Mentor

NVIDIA Corporation

Siddha Ganju is a Researcher and Data Scientist at NVIDIA, focusing on computer vision optimization for vehicle autonomy and medical instruments. She serves as a mentor on the Learn-to-Race affiliated capstone projects, in the Master's in Computational Data Science (MCDS) program, from the Language Technologies Institute, in the School of Computer Science, at Carnegie Mellon University. She is a 2016 MCDS alumna.

Core Team

Kevin Chian

Master's Student

CMU MCDS ‘23

Computer Vision developer working on integrating the latest fast and generalizable models.

Sidharth Kathpal

Master's Student

CMU MCDS ‘23

Reinforcement Learning algorithms developer building model based agents, and custom reward functions.

Arav Agarwal

Mater's Student

CMU MCDS ‘23

Reinforcement Learning algorithms development for the Safety policy agents.

Dhruv Arya

Master's Student

CMU MCDS ‘23

Computer Vision developer working on integrating techniques for distinguishing the track from surroundings.

Yujun Qin

Master's Student

CMU MCDS ‘23

Building the ROS stack for the Learn to Race team.

Tanay Gangey

Master's Student

CMU MCDS ‘23

Reinforcement Learning algorithms development for model based approaches.

Past Members

Ignacio Maronna Musetti

Saral Tayal

Alexey Skabelkin

Xinnan Du

Andrey Gostev

Jikai Lu

Sanil Pande

Christian Deverall

Abhinav Gupta

Zihang Zhang

Ruoxin Huang

Katik Chaudari

Sharvya Bhat

Jiayi Weng

Bhoui Fang

Poorva Agrawal

Technology

Racing Simulator

Partnered with Arrival to build a high-fidelity racing simulator, featuring software-in-the-loop and hardware-in-the-loop simulation capabilities. The simulator has played a key role in bringing autonomous racing technology to real life in the Roborace series, the world’s first extreme competition of teams developing self-driving AISimulator used to host the first Learn-to-Race Autonomous Racing Virtual Challenge to teach an artificially intelligent agent how to race.This is some text inside of a div block.

A Multimodal Control Environment for Autonomous Racing

Safe Autonomous Racing via Approximate Reachability on Ego-vision