Jun-Gill Kang
I'm a researcher at Defense AI Center of the Agency for Defense Development (ADD) in South Korea. I completed my undegraduated studies in POSTECH , South Korea advised by Prof. Soohee Han . I also worked with Prof. Hae-Won Park as visiting researcher in KAIST DRCD Lab.
I've selected as graduate representative and gave Commencement Speeches in POSTECH at 2022. Also my research youtube channel has >13K subscriber with >8M views total.
I have built and controlled a wall-climbing robot, a flying-squirrel-inspired drone, a quadrupedal robot, an off-road self-driving car, and an autonomous tunnel exploration robot.
I am currently seeking a PhD opportunity for Fall 2026 to further pursue research in Robotics and AI. If you believe my background and interests align with your program or lab, I would be glad to connect.
Email /
CV /
Scholar /
Youtube
Research
I'm interested in robotics, control, reinforcement learning, and deep learning. Most of my research start from building robot, fixing it and controlling it. Representative papers are highlighted . (Equal contributions are denoted by *)
Your browser does not support the video tag.
A Highly Maneuverable Flying Squirrel Drone with agility-improving Foldable Wings
Dohyeon Lee* ,
Jun-Gill Kang* ,
Soohee Han
IEEE Robotics and Automation Letters (RA-L) , 2025
project page
/
video
/
arxiv
Develop general trajectory tracking controller of flying squirrel with learned wing dynamics.
Fast, Perceptive Quadrupedal Locomotion in Complex Terrain
Jun-Gill Kang ,
JaeHyun Park,
Tae-Gyu Song,
Hae-Won Park
IEEE International Conference on Robotics and Automation Workshop (ICRA) , 2024
project page
/
video
/
paper
Control of full sized quadruped robot with vision pipeline.
Your browser does not support the video tag.
A Highly Maneuverable Flying Squirrel Drone with Controllable Foldable Wings
Jun-Gill Kang* ,
Dohyeon Lee* ,
Soohee Han
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2023
project page
/
video
/
arxiv
Build and control non-linear dynamics governed flying squirrel inspired quadrotor using reinforcement learning.
Your browser does not support the video tag.
Development of Dual-Unit Ceiling Adhesion Robot System With Passive Hinge for Obstacle Traversal Under Kinodynamic Constraints
Young-Woon Song,
Jun-Gill Kang ,
Son-Cheol Yu
IEEE Access , 2023  
project page
/
video
/
paper
Developed dual-unit wall climbing robot that can attach to fixed wall and traverse obstacle using both body.
Service
Reviewer for IROS2024, IROS2025, CORL2025