cv

Ph.D. Candidate

University of Michigan, Ann Arbor, MI

Robotics

• Robot Kinematics & Dynamics
• Nonlinear Systems
• Motion Planning
• Applied CUDA Programming
• Machine Learning

B.S.E.

University of Michigan, Ann Arbor, MI

Computer Engineering

• Minor in Mathmatics

B.S.E.

Shanghai Jiaotong University, Shanghai, China

Electrical & Computer Engineering

John Wu & Jane Sun Outstanding Scholarship

John Wu & Jane Sun Endowment Fund

The Scholarships for Excellent Students and Talents reward freshmen with excellent scores on the Entrance Examination as well as meeting favorable conditions

Jointed Tails Enhance Control of Three-dimensional Body Rotation

Royal Society Interface

Tails used as inertial appendages induce body rotations of animals and robots, a phenomenon that is governed largely by the ratio of the body and tail moments of inertia. However, vertebrate tails have more degrees of freedom (e.g., number of joints, rotational axes) than most current theoretical models and robotic tails. To understand how morphology affects inertial appendage function, we developed an optimization-based approach that finds the maximally effective tail trajectory and measures error from a target trajectory. For tails of equal total length and mass, increasing the number of equal-length joints increased the complexity of maximally effective tail motions. When we optimized the relative lengths of tail bones while keeping the total tail length, mass, and number of joints the same, this optimization-based approach found that the lengths match the pattern found in the tail bones of mammals specialized for inertial maneuvering. In both experiments, adding joints enhanced the performance of the inertial appendage, but with diminishing returns, largely due to the total control effort constraint. This optimization-based simulation can compare the maximum performance of diverse inertial appendages that dynamically vary in moment of inertia in 3D space, predict inertial capabilities from skeletal data, and inform the design of robotic inertial appendages.

Rapid and Robust Trajectory Optimization for Humanoid

arxiv

Under Review in IEEE ICRA 2025.

System Identification For Constrained Robots

arxiv

Under Review in IEEE Robotics and Automation Letters.

Safe Planning for Articulated Robots Using Reachability-based Obstacle Avoidance With Spheres

Robotics: Science and Systems

This paper introduces SPARROWS, a receding-horizon trajectory planner that generates real-time, provably-safe motion plans for articulated robots in unstructured environments. By using a novel reachable set representation and exact signed distance function, SPARROWS ensures collision-free trajectories while outperforming state-of-the-art methods in complex environments.

Can't Touch This: Real-Time, Safe Motion Planning and Control for Manipulators Under Uncertainty

arxiv

Under review in IEEE Transactions on Robotics.

Serving Time: Real-Time, Safe Motion Planning and Control for Manipulation of Unsecured Objects

IEEE Robotics and Automation Letters

This paper presents WAITR, a motion planning and control framework designed to safely manipulate unsecured objects with serial link manipulators in uncertain environments. Using reachability analysis, WAITR generates real-time, provably-safe motions, outperforming existing methods in both simulations and real-world experiments.

Reachable Sets for Safe, Real-Time Manipulator Trajectory Design

Robotics: Science and Systems

This paper presents ARMTD, a real-time trajectory planner for robotic arms that guarantees safety by incorporating reachable sets and fail-safe maneuvers in its motion planning. ARMTD generates provably-safe, collision-free plans and outperforms CHOMP in both simulations and real-world tasks, ensuring reliable operation in dynamic environments.

Safe, Optimal, Real-time Trajectory Planning with a Parallel Constrained Bernstein Algorithm

IEEE Transactions on Robotics

This paper introduces RTD*, an enhanced planning method for mobile robots that guarantees globally optimal, dynamically feasible plans at each iteration using a new Parallelized Constrained Bernstein Algorithm (PCBA). RTD* improves over the original RTD by addressing issues with liveness and optimality, demonstrating better performance in real-time planning tasks in complex environments with obstacles.

Rapid Offline Gait Optimization For Humanoids

An optimization toolbox to generate a library of physically feasible while energy efficient multiple-step gaits including different step lengths and step heights.

• Developed an optimization algorithm to generate a library of physically feasible while energy efficient multiple-step gaits including different step lengths and step heights.
• Parameterized walking gaits as Bezier curves to generate smooth trajectories.
• Efficiently handled closed-loop linkage on humanoids' ankles for controlling the orientation of the feet as differentiable kinematics constraints for faster optimization.
• Implemented the algorithm fully in C++ that outperforms all related works in terms of computation time and energy consumption.

Provably-safe Real-time Trajectory Optimization For Robotic Manipulators

A real-time optimization-based planning framework to generate provably safe trajectories, incorporating controller tracking errors.

• Developed a real-time optimization algorithm to generate provably safe trajectories, incorporating controller tracking errors.
• Enforced safety over continuous time intervals instead of discrete time instances for collision avoidance, robot torque limits, and gripper contact forces.
• Optimized constraint evaluation via parallel computation to significantly reduce computation time.
• Achieved higher collision avoidance success rates compared to existing methods.

Robust Control For Humanoids and Robotic Manipulators

A model-based robust controller that provides guaranteed uniform tracking error bound in presence of bounded robot model uncertainty.

• Construct the tracking error bound by constraining the upper bound of a Lyapunov function.
• Incorporated closed-loop linkage dynamics on humanoids' ankles as kinematics constraints into the controller for better tracking performance on humanoids.

System Identification For Humanoids and Robotic Manipulators

A general optimization-based system identification algorithm for both humanoids and robotic manipulators to estimate robot inertial parameters and motor friction parameters.

• Provided bounds on estimation of the robot model parameters based on the Cramér–Rao bound to integrate it into the robust controller for guaranteed tracking performance.
• Implemented the whole pipeline in C++ for faster computation time.