Jeongbin Yun > Internship

본문 바로가기
Y
CURRENT INTERN

Jeongbin Yun 윤정빈

VR-Based Humanoid Robot Teleoperation · Humanoid Whole-Body Control · Robot Kinematics and Inverse Kinematics · Human–Robot Motion Mapping and Motion Retargeting · Trajectory Generation and Tracking · PD Control and Controller Tuning · Robot Simulation with MuJoCo · Real-Time Robot Control · ROS2-Based Robot Systems
yjb1018@naver.com 989-111 Daedeok-daero, Yuseong-gu, Daejeon 305-353, Korea
01 PROFILE

Project Role and Responsibilities

My work focuses on VR-based humanoid robot teleoperation and robot control, with particular interest in transferring human motion from virtual reality interfaces to humanoid robotic systems while maintaining accurate and stable robot motion.

My main responsibilities include:

  • Developing VR-based teleoperation systems using Unity and Meta Quest
  • Implementing real-time communication between VR interfaces, robot controllers, and simulation environments
  • Developing inverse-kinematics and whole-body control methods for the Unitree G1 humanoid robot
  • Analyzing coordinate transformations, joint configurations, reachability, and robot motion constraints
  • Developing and evaluating robot motion trajectories in MuJoCo simulation
  • Tuning joint-level control parameters for accurate and stable trajectory tracking
  • Integrating simulation-based control methods with physical humanoid robot operation

I primarily work with Unity/C#, Python, ROS2, MuJoCo, Mink, Meta Quest, and Unitree G1 robot control interfaces.

Project Contributions

Unitree G1 Humanoid Teleoperation
Developed and evaluated a VR-based teleoperation pipeline for the Unitree G1 humanoid robot, transferring hand position and orientation data acquired from Meta Quest through Unity to robot simulation and control environments.

Whole-Body Inverse Kinematics and Motion Control
Implemented and evaluated whole-body inverse-kinematics-based control using MuJoCo and Mink. Investigated robot joint configurations, redundancy, reachability, coordinate transformations, and motion constraints to improve end-effector tracking performance.

Robot Motion and Controller Optimization
Designed repeated motion and trajectory tests to analyze humanoid joint behavior and tracking performance. Evaluated and tuned PD controller parameters to improve motion accuracy, responsiveness, and stability during bidirectional and continuous robot motion.

Real-Time VR–Robot Integration
Developed communication and control pipelines connecting VR input, simulation, and humanoid robot control. Investigated discrepancies between simulated and physical robot behavior and contributed to improving the transition from simulation-based control to real Unitree G1 operation.

02 EDUCATION
Mar. 2021 – PresentB.S. Student in Robotics Engineering, Hanyang University ERICA, Republic of Korea
03 RESEARCH INTEREST
VR-Based Humanoid Robot TeleoperationHumanoid Whole-Body ControlRobot Kinematics and Inverse KinematicsHuman–Robot Motion Mapping and Motion RetargetingTrajectory Generation and TrackingPD Control and Controller TuningRobot Simulation with MuJoCoReal-Time Robot ControlROS2-Based Robot Systems