Career History

Thesis: Accelerating Policy Learning for Robust Control of Robotic Manipulators and Aerial Vehicles via Physics-Informed Guidance

​Supervisor: Professor Dong Eui Chang

Experience Summary: Recipient of KAIST scholarship. Throughout my MS program, I also worked as a graduate research assistant at Control Lab under my supervisor. I specialized in the intersection of classical control theory and learning-based algorithms for complex robotic systems. ​My work spanned theoretical studies and designs to real-world hardware deployment on manipulators, drones, and mobile robots.

​

Primary Research: 

• Marine-Manipulator Control (Hanwha Aerospace): I was a part of the research project involving control of a marine-manipulator system under adverse sea conditions. I played key roles on all major aspects of the project since the inception, including simulation design, system identification, hardware interface, algorithm verification, and the control system design, implementation, training and optimization. I have a co-authored paper prepared to be submitted.

• Robust Drone Control: I performed independent research on robust drone control strategies that combines model-based and geometric control strategies to reinforcement learning algorithms for robust control under wind and swinging payloads. I have a first author paper under review at IEEE RAL and a co-authored paper being prepared.

• Hybrid Control: I also did collaborative research on robot independent hybrid control strategies utilizing model-based approaches to improve learning-based solutions for sample efficiency in  RL (model-based baselines, approximate analytic solutions, residual RL), parameter estimation and tuning (using RL, differentiable simulations, and observer designs), and on reducing sim-to-real gap for RL policies (robust simulation design, disturbance observers, L1-adaptation, and offline fine-tuning approaches).

​

Technical Expertise:

• Control Theory: Nonlinear control  (Lyapunov analysis, observer designs both control theoretic like super-twisting and stochastic like Kalman filters,nand nonlinear control designs), Optimal control (LQR, MPC, Approximate HJB), and geometric control for manifolds (think rigid body rotations).

• Learning and AI: Deep RL, Foundation models (VLAs, LLMs), World models (RSSM/Dreamer), and Neural ODEs for system identification.

• Differentiable Simulations: Nvidia Warp/Newton, JAX and Pytorch for differentiable simulation based policy learning and system identification.

• Manipulator Control: Control strategies specifically for manipulators including FK-IK computations, trajectory designs, observer designs, and system modeling through screw-axis approach (refer to Lynch-Park Modern Robotics).

• Hardware Implementations: High-performance control loops in C/C++ on embedded systems, TensorRT and CUDA implementations on edge devices like Jetson for RL policies.

​

Some other very niche skills I learned or experienced (if they are of interest to you): 

• Designing simulation models, either from scratch (first-principles approach), or through tools like Nvidia IsaacSim, Warp, JAX, MATLAB and Simulink.

• Implementation and optimization of existing research work.

• Porting of existing work between Python/MATLAB/C++. 

• Integration schemes for discrete-time implementation of control algorithms and for simulations.

• Dealing with very nonlinear and badly saturated actuators, and crazyflie drones.