My research applies AI robotics technology on autonomous race cars.

Autonomous racing presents unique challenges and opportunities to general AV research.

Optimal Control

Autonomous race cars should, as their human counterparts, learn about the dynamics of the vehicle and track conditions on the fly, and optimize over laps. I research in learning-based model-predictive controller that combines ML techniques and optimal control for safety-critical systems with unknown changing dynamics.

AI Robotics

RL and foundation models has the potential to achieve more advanved robot autonomy. The reliable evaluation of such policy on real-time safety-critical systems could bring innovation to AV and robotics systems in extreme environments.

Open Source Robotics

Overtime, the open-source community around ROS and ROS2 has made open AV architecture readily available. Autonomous racing, however, pushes roboticists to re-think about these stacks. I am an active contributor to the open source robotics world, with contributions ranging from MPC to sensor drivers.

I lead the best U.S. team in the Indy Autonomous Challenge

๐Ÿฅˆ 2nd Place, Indy Autonomous Challenge at Texas Motor Speedway 2022

๐Ÿฅ‰ 3rd place, Indy Autonomous Challenge at Las Vegas Motor Speedway 2023

I’m interested in:

  • Learning Model Predictive Control (LMPC) for multi-car autonomous racing.
  • Trajectory optimization for autonomous racing.
  • Multi-modal perception pipeline for autonomous racing: 3D detection (Lidar CenterPoint), 2D detection (Yolo), object fusion, EKF-based vehicle tracking.
  • RL and LLM combining high-level planning tasks and low-level reflexes

I have full-stack experience in autonomous driving, from high-level autonomy software architectures to drive-by-wire system programs. I deploy full-size, safety-critical robotics systems in the field.

Learn More About My Projects

Publications

Learning Model Predictive Control with Error Dynamics Regression for Autonomous Racing
Haoru Xue, Edward L. Zhu, Francesco Borrelli
Submitted to ICRA 2024 | arxiv | video | code

Spline-Based Minimum-Curvature Trajectory Optimization for Autonomous Racing
Haoru Xue, Tianwei Yue, John M. Dolan
Submitted to ICRA 2024 | arxiv | code

Model Predictive Control for Integrated Lateral Stability
Jad Yahya, Siddharth Saha, Haoru Xue, Allen Y. Yang
Submitted to IEEE TAC | arxiv | code

Contacts

Find me on GitHub, LinkedIn, or reach out in email. I also have a YouTube channel for some interesting research clips in autonomous racing.
Haoru Xue

Haoru Xue

M.S. Robotics, Carnegie Mellon

Find me on GitHub, LinkedIn, or reach out on email. I also have a YouTube channel for some interesting research clips in autonomous racing.