Dohyeon Lee

I am an undergraduate student in the Department of IT Convergence Engineering at POSTECH in South Korea. I am currently serving in the Republic of Korea Army to fulfill my mandatory military duty (until December 2025).

Email  /  Scholar  /  Youtube  /  CV

Research

I'm interested in robotics, control, motion planning, deep learning. Representative papers are highlighted.
(Equal contributions are denoted by *)

Develop general trajectory tracking controller of flying squirrel with learned wing dynamics.

Build and control non-linear dynamics governed flying squirrel inspired quadrotor using reinforcement learning.

Project

• Environment estimation from the sample history of MPPI.
• Obtain the reachable set using an empirical gramian based method.

• Implement integral backstepping based Lyapunov stable controller for drone.
• Implement adaptive Kalman filters for state estimation under aggressive maneuvers.
• Collect data using Mocap and GNSS for training neural net dynamics model.
• Implement simulation for full SE(3) dynamics model using MATLAB using ode45 integrator.

Skills : Linkage system design, 3D CAD, Hardware design, GNSS, Mocap, Learning from demonstration

• Development of 6-axis manipulator hardware and stabilization of low-level control system.
• Proposal and Implementation of a Physics-Informed RNN for water state observer.
• Implement LQR controller and velocity motion planning of manipulator for water sloshing control.

Skills : 3D CAD, 3D printing, Physics-Informed Learning, ROS2

• implement off-policy (SAC) and on-policy (PPO) RL to a Legged-manipulator robot in the Isaac Gym.
• Designed a learning pipeline for whole-body control of a legged-manipulator robot using a VAE and replay buffer of an off-policy RL.

Skills : OSC(operational space controller), Isaac Gym, RL

• Develop Glidable drone with foldable wings.
• Simulating gliding drone with matlab.

Skills : 3D CAD, 3D printing, Kalman Filter

• Development of quadruped-wheel robot hardware and stabilization of low-level control system.
• Balancing control with inverse kinematics.
• Application of impedance control to the robot legs.

Skills : 3D CAD, 3D printing, CAN protocol, UDP protocol, simulating impedance control with matlab

• Directly implement mapping and A* algorithms with Python.
• Development SLAM algorithm from scratch.
• Hardware development.

Skills : 3D CAD, 3D printing, ROS1

In the Media

About Flying Squirrel Drone Research
International Media Coverage: TechXplore(Cover Article) , NEW ATLAS(Cover Article) , heise online , TechieTonics(Personal Interview), ...
Domestic Media Coverage: SBS(Korea Major Media) , YTN(Korea Major Media) , ...

template