About Me

I am an NSF Graduate Research Fellow in the General Robotics Lab at Duke University, advised by Boyuan Chen. Prior to joining the GRL, I was a member of the Dynamical Systems Lab advised by Brian Mann.

I research robotics, dynamical systems/mechanics, learning, and optimization. Recently, I've been focused on developing smooth neural dynamics for model-based reinforcement learning and whole-body control. I also like physiology and biomechanics, which I studied as an undergrad.

I like working on challenging problems that sit at the intersection of multiple fields, and I enjoy using accelerated computing, differentiable programming, and numerical methods to tackle these problems.

In my free time, I enjoy watching Cincinnati sports (regrettably), calisthenics/weight lifting, all things outdoors, and just hanging out with my cat, Molly. I might also be one of the only valid Kentucky and Duke fans out there.

I am actively seeking a full-time position in industry or academia after my graduation in summer 2026. Please feel free to contact me if my background aligns with your interests.

Research

Learning Legged MPC with Smooth Neural Surrogates [Project page] [Paper] [Code] [YouTube] We introduce a neural dynamics model that is scalable, generalizes reliably, and excels at gradient-based optimization at test time. The result is the smooth neural surrogate: an architecture with explicit smoothness control to attenuate stiff contact transitions and interpolate reliably in regions where data are sparse or standard neural networks fail.

Sym2Real: Symbolic Dynamics with Residual Learning for Data-Efficient Adaptive Control [Project page] [Paper] [Code] [YouTube] We present a data-driven framework for training low-level adaptive controllers in a highly data-efficient manner. Using only about 10 trajectories, we achieve robust control of both a quadrotor and a racecar in the real world, without expert knowledge or simulation tuning.

Automated Global Analysis of Experimental Dynamics through Low-Dimensional Linear Embeddings [Project page] [Paper] [Code] [YouTube] This work presents an end-to-end approach for discovering hidden stuctures in experimental data from dynamical systems. We simulaneously address challenges in applied dynamical systems due to mathematical modelling, nonlinearity, and high dimensionality through low-dimensional linear embeddings of nonlinear dynamics.

LAPP: Large Language Model Feedback for Preference-Driven Reinforcement Learning [Project page] [Paper] [Code] [YouTube] Here, we use LLMs to replace human feedback in preference-based reinforcement learning. Our method, Large Language Model Assisted Preference Prediction, leverages LLMs to generate feedback on agent behaviors, enabling efficient policy learning, and behavior shaping without explicit reward engineering.

Semantic Visual Self-Modeling for Whole-Body Awareness in Legged Robots [Code] In this project, we developed learned 3D models of a legged robot's body to instill whole-body awareness for downstream planning or world models. I hope to soon combine this work with my ongoing research on smooth neural surrogates, and contact-rich control.

Differentiable Direct Collocation for Hardware-Software Co-Design [Code] This project aimed to develop methods to understand how a system’s physical parameters influence its controllability, stability, maneuverability, etc.. I did this by differentiating through a closed-loop controller around a nominal trajectory obtained via trajectory optimization. I developed a differentiable direct collocation library in JAX via IPOPT and implicit differentiation, then obtained some preliminary results on a simple cart-pole.

Comparison of Data-Driven Methods on Discovering the Dynamics of the Unforced Multi-axis Cart System [Paper] In this work, we compared various data-driven methods for uncovering the dynamics of the unforced multi-axis cart system. This system is particularly interesting as it can be configured as either a linear or nonlinear system, which enable us to evaluate the strengths and weaknesses of each method in both regimes.

Stability Prediction via Parameter Estimation from Milling Time Series [Paper] We introduce an approach to automated milling stability analysis whereby the vibration behavior of a milling tool during cutting is used to directly obtain a bifurcation diagram of the system in real-time.

A Model-Free Sampling Method for Basins of Attraction Using Hybrid Active Learning [Project page] [Paper] This work explores hybrid active learning as an efficient method to automatically discover the basins of attraction of dynamical systems.

Supervised Learning for Abrupt Change Detection in a Driven Eccentric Wheel [Paper] This paper compares supervised learning methods for phase identification (i.e., roll, slip, and hop) in simulated data from a driven eccentric wheel as a prototypical non-smooth system.

The Eccentric Disk and Its Eccentric Behavior [Paper] An eccentric disk has a non-constant normal force and therefore has four distinct phases of motion: oscillations about a stable equilibrium, roll without slip, roll with slip, and hop. In this work, the system is analytically modeled using an augmented Lagrangian formulation, solved with numerical integration, and experimentally realized.

Show more

Education

	Duke University - Durham, NC, USA 2026 (expected), Ph.D. in Mechanical Engineering and Materials Science
	University of Kentucky - Lexington, KY, USA 2019, B.S. in Exercise Science and Statistics (Topical Studies)