my photo

Joonseo Kym

Undergraduate Student
@ Dep. of Mechanical Engineering
& Electriclal Engineering, KAIST

I am dedicated to advancing highly dynamic robotics and intelligent control systems. While my recent research involves developing optimal pneumatic mechanomyography (pMMG) sensors and real-time signal processing pipelines, my core objective is to bridge hardware design and learning-based control to build robust robotic systems.

I possess a solid engineering skillset that includes mechanical design (CAD), real-time data acquisition, and signal processing. Leveraging my dual background in mechanical and electrical engineering, I have a strong foundation in system modeling and control theory. I am highly motivated to deepen my expertise in custom electronics and PCB design to develop fully integrated robotic hardware. With a growing interest in reinforcement learning (RL), I am eager to contribute to cutting-edge research in legged locomotion and autonomous robotics.

Contact me Download CV (PDF)

Education

KAIST (Korea Advanced Institute of Science and Technology)

Bachelor of Mechanical Engineering & Electrical Engineering • Expected Jul 2029

GPA: 3.77 / 4.30

Relevant Coursework: Introduction to Robotics Engineering, System Modeling and Control, Mechatronics System Design, Mechanical Vibrations, Data Structures, Solid Mechanics, Dynamics, Signals and Systems, Circuit Theory

IASA (Incheon Academy of Science and Arts)

National School for Gifted Students • Mar 2019 — Feb 2022

Research Experience

Exoskeleton Lab (KAIST)

Undergraduate Research Assistant (Individual Research & URP) • Dec 2023 — Jan 2025

  • Conducted comprehensive research on proportional Mechanomyography (pMMG), encompassing theoretical principles, analog pneumatic sensor selection, and the end-to-end hardware manufacturing process under Professor Kyoungchul Kong.
  • Fabricated custom pMMG sensors and developed a robust data acquisition (DAQ) system utilizing NI myRIO and FPGA architecture.
  • Validated signal integrity and performance metrics through low-level LabVIEW programming, ensuring precise sensor calibration for robotic exoskeleton applications.
  • Prepared a full manuscript as the 2nd author, detailing the design, fabrication, and validation methodologies of the customized pMMG sensor system.

Patents & Publications

Elbow Rehabilitation Aiding Device View Details

Registered Patent (KR App No. 10-2021-0034782) • Feb, 2024

Co-inventor. Developed a novel elbow rehabilitation device utilizing a string twisting actuation mechanism. Collaborated with Daeki Cho, Inwoo Kim, Jaewoo Park, Doyoon Kim, and Seokjoo Choi at Incheon Academy of Science and Arts.

Awards & Honors

Fourth Award (Biomedical Engineering) & Science Champion Award

Regeneron ISEF & USAID • May 2021

  • Recognized as one of the top research projects globally in the world's largest pre-college science competition (ISEF).
  • Awarded the Science Champion Award by the U.S. Agency for International Development (USAID) for innovative biomedical contributions.

Gold Prize (Ranked 1st in Physics/Earth Science)

Samsung Electronics (Samsung Humantech Paper Award) • Feb 2021

Received the highest honor in the Physics/Earth Science category at South Korea's most prestigious corporate-led research competition (27th Samsung Human-Tech Thesis).

Grand Prize (Minister of Science and ICT Award) & Gold Prize

Ministry of Science and ICT (YSC) & KSEF • Nov 2020 — Feb 2021

  • Awarded the Grand Prize by the Ministry of Science and ICT (YSC) for excellence in innovative biomedical research.
  • Achieved the Gold Prize at the 18th Korea Science and Engineering Fair (KSEF).

Research Excellence Award (Honorable Mention)

Korea Science Foundation • Feb 2020

Recognized for outstanding biomimetic research at the national level for the project "Development of a Landing Device Inspired by Cat Skeletal Structure."

Research Projects

Optimal pMMG Sensor Placement for Knee Extensor Analysis

KAIST URP & Follow-up Research • Project Lead & First Author • Jun 2024 — Present

  • Independently conceptualized the research topic and spearheaded the entire project lifecycle—from initial hypothesis formulation to hardware design and experimental validation.
  • Developed a synchronized data acquisition system using an STM32 microcontroller and MS5607 pressure sensors to capture muscle volume changes via pneumatic mechanomyography (pMMG), overcoming traditional sEMG/IMU limitations.
  • Processed kinematic data using gait phase segmentation and Dynamic Time Warping (DTW) to map pMMG signals, establishing optimal sensor placement guidelines (e.g., 20% muscle length for vastus lateralis).
  • Designed and executed advanced follow-up experiments at Angel Robotics using Motion Capture systems and Force Plates to simultaneously measure pMMG signals and actual knee joint torque via inverse dynamics.
  • Currently drafting a first-author manuscript for journal submission detailing the quantitative assessment of knee extensor activation and joint torque estimation using pMMG sensors.

Development of a Circular Twisted String Continuous Passive Motion Actuator

IASA SA (Steam Activity) Research Project • First Author / Project Lead • Mar 2020 — Apr 2021

  • Developed a wearable, circular twisted string continuous passive motion (CPM) actuator to significantly reduce the volume and weight of traditional elbow rehabilitation devices.
  • Engineered a 2-stage pulley system within the circular mechanism, which quadrupled the range of motion and speed to achieve a 10.6 cm operational range sufficient for rehabilitation.
  • Established and experimentally validated a mathematical model to predict string displacement and output force based on the number of twists.
  • Demonstrated through load testing with linear potentiometers that the actuator generates over 60N of force, surpassing the 32.6N benchmark required to move an adult male forearm at a constant velocity.
  • Achieved up to a 15-fold increase in output force relative to motor torque, confirming the system's mechanical stability and efficiency for high-load rehabilitation tasks.

Automatic Shuttlecock Classification & Arrangement Sweeper

ME340 Engineering Design Project • Student Engineer • Sep 2023 — Dec 2023

  • Designed and prototyped a mechanical sweeper incorporating a wheel-driven conveyor belt and a 3-pulley mechanism to automatically collect and lift badminton shuttlecocks from the ground.
  • Engineered a smart classification module using an Arduino and 16 IR light sensors to evaluate feather integrity by analyzing light transmission and standard deviation, automatically separating damaged units.
  • Fabricated the hardware utilizing SolidWorks, aluminum profiles, and 3D-printed components, while optimizing gear ratios (12/225) for efficient conveying velocity and integrating brush-type propellers to prevent payload damage.
  • Validated the system's efficacy through focus groups, A/B testing, and competitive analysis against commercial products, demonstrating superior performance in cost-effectiveness and user convenience.

Optimization of Deep Reinforcement Learning for Control Problems

KAIST Pre-URP Program • Student Researcher • Aug 2019 — Sep 2019

  • Implemented and trained a Double Deep Q-Network (DDQN) agent to solve the CartPole control problem using Keras and OpenAI Gym environments.
  • Conducted extensive hyperparameter tuning across 11 variables (e.g., learning rate, epsilon decay, network depth, activation functions) to stabilize the learning process and maximize reward acquisition.
  • Evaluated the model's performance over 1,000 episodes, establishing that the optimized DDQN model achieved the success threshold (average score > 195) in just 151.8 episodes on average.
  • Significantly improved learning efficiency, outperforming the default model's baseline of 549.6 episodes by approximately 72%.
  • Completed intensive 2-week on-site research mentored by Professor Jeha Ryu at the Gwangju Institute of Science and Technology (GIST).

Development of a Drone Landing Gear Mimicking Feline Skeletal Structure

IASA SA (Steam Activity) Research Project • Student Researcher • Mar 2019 — Dec 2019

  • Designed and 3D-printed a biomimetic drone landing gear inspired by a cat's skeletal structure to ensure safe landings and minimize payload damage.
  • Engineered a shock-absorbing joint mechanism utilizing the repulsive force of neodymium magnets and PVC foam, completing 7 iterative design revisions to optimize joint angles up to 152 degrees.
  • Conducted drop experiments using PASCO ultrasonic sensors and Capstone software to analyze kinematics (position, velocity, acceleration), identifying that a radial leg arrangement yields superior buffer performance.
  • Validated the prototype's shock-absorbing capabilities, successfully achieving stable landings for a 1.5kg payload from a 1m height and a 3kg payload from a 50cm height.
Note: Publications and projects prior to 2021 were authored under the name 'Joonseo Kim'.

Technical Skills

Languages

Korean (Native), English (Fluent), German (Basic)

Programming & Software

Python, C, SolidWorks (CAD)

Robotics & Engineering

System Modeling & Control, Dynamics, Mechanical Vibrations, Circuit Theory

Hardware & Prototyping

Mechatronic System Integration, 3D Printing, Sensor Fabrication (pMMG), Actuator Interfacing

Research & Analytics

Real-time Data Acquisition, Signal Processing, Experimental Validation, Scientific Writing

Military Service

Gwacheon Family Center

Social Service Personnel (Mandatory Military Alternative Service) • Feb 2025 — Nov 2026

  • Developed a Python-based automated transportation cost calculator with a custom GUI, integrating the Kakao Address API to streamline complex reimbursements, reducing manual administrative workload by 95% and improving data accuracy.
  • Completed mandatory military alternative service by providing educational mentorship to children from multicultural families and supporting community engagement initiatives.
  • Managed official document processing and assisted in daily operational activities for efficient center management.

Associations

KAIST Student Ambassador (KAINURI)

President (Official University Representative) • Mar 2022 — May 2024

Led the official student ambassador team, KAINURI, serving as a liaison between KAIST and the public to enhance the university's image.

  • Organized and participated in events to promote KAIST's initiatives and achievements to external audiences.
  • Developed and managed content for KAINURI's official social media platforms, increasing follower engagement by 25%.
  • Collaborated with various university departments to coordinate promotional activities and campus tours for prospective students.

KAIST Creative Global Leader Camp

Head Teaching Assistant & Camp Coordinator • Mar 2022 — May 2024

Coordinated a camp designed to foster creativity and global leadership among high school students.

  • Developed and facilitated interactive workshops and hands-on projects to enhance participants' scientific thinking and problem-solving skills.
  • Supervised a team of teaching assistants, ensuring the smooth execution of camp activities and providing mentorship to participants.
  • Managed logistics, scheduling, and communication with participants and their guardians, resulting in a 95% satisfaction rate among attendees.

Social Links

  • Github: https://github.com/joonseokym
  • LinkedIn: https://www.linkedin.com/in/joonseokym/
  • Instagram: https://www.instagram.com/ssemiwest/