Hansol Park
Curriculum Vitae · AI Robotics Engineer
M.S. student in electrical, electronics and communication engineering, researching hierarchical ACT / RL for dexterous manipulation and working toward reactive dexterous manipulation.
Education
M.S. in Electrical, Electronics & Communication Engineering
Korea University of Technology and Education (KOREATECH) · Cheonan, South Korea
Advisor: Prof. Yong-Jae Kim
Current research: Hierarchical ACT / RL for dexterous manipulation.
Research interest: Reactive dexterous manipulation.
Expected Feb. 2028
B.S. in Mechatronics Engineering
Korea University of Technology and Education (KOREATECH) · Cheonan, South Korea
Laboratory & capstone advisor: Prof. Hongkeun Kim
Capstone: SYSID Sim-to-Real Using a Two-Wheeled Legged Robot
Academic foundation: Control systems; reinforcement learning and sim-to-real; mechanical design; electronics and embedded systems.
Research Interests
- Reactive Dexterous Manipulation
- Hierarchical ACT / RL Policies
- Robot Learning
- Vision-Language-Action Models
- Teleoperation & Robot Data
- Sim-to-Real Deployment
Research Experience
M.S. Student Researcher
IRIM Laboratory, KOREATECH · Cheonan, South Korea
Studies hierarchical ACT / RL policies for dexterous manipulation on the 22-DoF ALLEX arm-hand platform, with a broader research interest in reactive dexterous manipulation.
- Participates in WIRobotics AI Team research as an IRIM Laboratory student researcher.
- Collected and curated approximately 85 teleoperated demonstrations at 50 Hz, prepared a LeRobot dataset, and trained an ACT policy for real-robot deployment.
- Validated learned policies in Isaac Sim before deployment on the physical robot.
Research Intern
IRIM Laboratory, KOREATECH · Cheonan, South Korea
Completed a pre-graduate research internship on robot-learning infrastructure for the ALLEX platform.
- Developed and organized ROS 2 control and state interfaces for manipulation experiments.
- Connected simulation, demonstration collection, dataset preparation, and policy-evaluation workflows.
Conference Presentations & Demonstrations
SYSID-A-RMA Sim-to-Real for a Two-Wheeled Legged Robot
Hansol Park, Hyeokjo Kwon and Seongwoo Bae
KRoC 2026 RED Show · Korea Robotics Society · Pyeongchang, South Korea
Teaching Experience
Control Engineering
School of Mechatronics Engineering, KOREATECH
Microprocessors and Laboratory
School of Mechatronics Engineering, KOREATECH
Selected Research Projects
ACT Training and Real-Robot Deployment on ALLEX
An end-to-end imitation-learning study for a 22-DoF ALLEX arm-hand system, from 50 Hz demonstrations to ACT training and real-robot task execution.
- Collected and curated approximately 85 teleoperated demonstrations at 50 Hz for a 7-DoF arm and 15-DoF hand.
- Converted raw demonstrations into a LeRobot dataset and trained an ACT policy for 1,000,000 steps on an RTX 5080.
Python · ROS 2 · LeRobot · ACT · Isaac Sim · PyTorch
Teleoperation and Robot Learning Data Pipeline
A compact public overview of the interfaces used to operate ALLEX, observe robot state, and prepare demonstrations for learning experiments.
- Iterated on a 7-DoF master-device interface and operator workflow for ALLEX-oriented teleoperation.
- Connected ROS 2 state streams, live robot visualization, and simulation/real-robot validation paths.
ROS 2 · Python · Isaac Sim
RedShow SYSID-A-RMA Sim-to-Real Project
A wheel-legged robot project that combined system identification with an A-RMA-style adaptation pipeline to reduce the simulation-to-reality gap.
- Built a reinforcement-learning workflow for a six-motor wheel-legged robot with position-controlled legs and velocity-controlled wheels.
- Combined identified physical parameters with an adaptation module instead of relying only on broad domain randomization.
Python · IsaacLab · RSL-RL · OpenCR · CAN · RS-485
ILRC 2026 Logistics Robot Sorting
Team KAISER's logistics sorting robot for the 2026 International Logistics Robot Competition, awarded first place and the Minister of Employment and Labor Award.
- Contributed as a Team KAISER member to integration and competition testing for the logistics sorting system.
- Supported a workflow combining stable robot control with vision-based object classification.
ROS 2 · Computer Vision · Robot Control
Fellowships & Awards
2026 ILRC · 1st Prize
Minister of Employment and Labor Award · International Logistics Robot Competition
First place in the university Logistics Robot 2 · Sorting division, recognized for stable robot control and accurate vision-based logistics classification.
WIRobotics · AWS / NVIDIA Physical AI Fellowship
2026 Physical AI Fellowship Cohort · MassRobotics · AWS · NVIDIA
WIRobotics joined the nine-company 2026 cohort to advance deployable physical AI and robot-learning workflows.
K-Community Excellent Team
Excellent Team Scholarship · KOREATECH
Team recognition after a winter robotics learning program covering Python, Arduino, ROS, OpenCR, and TurtleBot bring-up.
Technical Skills
- Programming
- Python · C and C++ (working knowledge)
- Robot Software
- ROS 2 · Linux · LeRobot
- Robot Learning
- PyTorch · ACT · Reinforcement Learning · RSL-RL · Sim-to-Real
- Simulation
- Isaac Sim · Isaac Lab · MuJoCo · Newton
- Mechanical & Embedded
- SolidWorks · OpenCR · CAN · RS-485
Languages
- Korean
- Native
- English
- TOEIC 785