I'm a researcher at the Defense AI Center of the Agency for Defense Development (ADD) in South Korea. I completed my undergraduate studies at POSTECH , advised by Prof. Soohee Han . I also worked with Prof. Hae-Won Park as a visiting researcher in the KAIST DRCD Lab .
I've been selected as a graduate representative and gave commencement speeches at POSTECH in 2022 graduation ceremony. My research YouTube channel has >13k subscribers and >8.7M total views.
I believe that computational modeling of physics and scene understanding is key to developing robots with genuine physical understanding. Most of my research begins with: building real robots, developing perceptual systems that interpret the world, simulating their dynamics, and controlling them in the wild.
Representative papers are highlighted. (Equal contributions are denoted by *.)
Revisable denoising for robotic sequential prediction through structured re-noising and cross-horizon refinement.
A unified perceptive policy enables high-speed multi-skill locomotion, automatic gait selection, and stable skill transitions on complex terrain using only onboard perception and computation.
Task-Oriented Hash-structured Occupancy Grid Mapping for efficient planning.
Disentangling context vectors in meta-learning enables robust and generalizable regression and locomotion across diverse tasks.
Developing a general trajectory tracking controller for a flying squirrel drone by learning wing dynamics and coordinating thrust–wing control for agile maneuvers.
Vision-based control of a full-sized quadruped robot in complex outdoor terrain using a perception pipeline and learned locomotion policies.
Building and controlling a highly nonlinear flying-squirrel-inspired quadrotor using reinforcement learning to manage foldable wings and thrust in challenging maneuvers.
A dual-unit wall-climbing robot that can adhere to ceilings and traverse obstacles using a passive hinge under kinodynamic constraints.
06.2022~06.2026
Full-size quadruped · outdoor terrain · vision-based control.
Skills: Perception, Depth Camera, 2D LiDAR, Sim-to-real
Video
12.2024~06.2026
Implement LiDAR based navigation foundation model.
Skills: VLM, Large transformer training. Imitation learning.
Technical report ·
Video
08.2023~08.2024
Flexible voxel hash mapping for LiDAR map.
Skills: Lidar-inertial odometry (LIO). ROS2. LiDAR perception.
Video
08.2023~08.2024
Off-road traversal of self-driving car for military purpose.
Skills: Model predictive path integral (MPPI). Neural network dynamics. LiDAR perception.
08.2023~10.2025
Multi-robot exploration with limited communication band.
Skills: ROS2. Lidar-Inertial-Odometry Algorithms. Navigation & Exploration algorithms.
Video
06.2021~12.2021
Develop wheel-legged robot with CAN protocol.
Skills: 3D CAD. 3D printing. Simulating whole-body lagrangian dynamics. Sequential QP based motion planning.
Video
02.2022~06.2022
Develop glidable drone with foldabe wings.
Skills: 3D CAD. 3D printing. State estimation. Kalman filter.
Video
06.2020~06.2022
Ceiling adhesion robot with laser projector based inspection mission plannig.
Skills: 3D Printing. Physics modelling. STM32 board programming.
Project page ·
Video
09.2020~12.2020
Developped 3D printing engine & Arduino with E-CAD
Skills: 3D CAD & Rendering. E-CAD.
Video
09.2020~12.2020
Developped cycloidal motor reducer with ROS2 interface
Skills: 3D CAD & Rendering. ROS2.
Video
Informal interview from IEEE Spectrum (Evan Ackerman), New atlas (Cover article), Tech Xplore (Cover article), SBS (Korea major press), YTN (Korea major press), and many others.
Reviewer for IROS 2024, 2025, RA-L 2026, CoRL 2025, 2026.
