M9G | Research & Archive
COMPLETE KNOWLEDGE ARCHIVE · MINCASURONG.COM × GITHUB.IO

Patents, Publications, Medical Robotics & Agentic AI Milestones

The unified repository of every patent, research article, clinical rehabilitation robot video (SUBAR & AI Therapist), Gemini 3 Seoul Hackathon epilogue, Google Cloud OnBoard technical recap, Sogang University LQR/MATLAB control lecture, and global exchange milestone from mincasurong.com.

Patents (US & KR)
14 Patents
Exoskeletons, Torque & Gait AI
Peer-Reviewed
20 Papers
ICRA, IEEE TNSRE, ICORR (2016–2026)
Clinical & TA Media
21 Videos
14 Rehab/SUBAR + 7 LQR/MATLAB
Clinical Deployment
6 Hospitals
KFDA Certified Medical Robot SUBAR
PEER-REVIEWED RESEARCH · 20 PUBLICATIONS (IEEE ICRA, IEEE TNSRE, JKROS, BIOROB, ICORR, SCIE, ICCAS)

20 Research Publications — Frontier IEEE Journals & Flagship Conferences First (2016 – 2026)

Highlighting flagship first-author and peer-reviewed publications in IEEE ICRA 2025, IEEE Transactions on Neural Systems and Rehabilitation Engineering (TNSRE 2020 & 2019), JKROS 2026, IEEE BioRob 2020, IEEE ICORR 2017, and Applied Sciences (SCIE), followed by conference proceedings, preprints, and thesis.

ICRA 25
★ Flagship IEEE Conf · 1st AuthorOPPA: Online Planner's Parameter Adaptation for Enhanced Mobile Robot Navigation
Minsu Chang, Junwon Jang, Daewoong Han, Wonje Choi, Seungyeon Kim, Hyunkyu Park, Hyundo Choi
2025 IEEE International Conference on Robotics and Automation (ICRA), pp. 3861–3867, May 19–23, 2025, Atlanta, USA · DOI: 10.1109/ICRA55743.2025.11128768
IEEE Xplore ↗
TNSRE 20
★ Frontier IEEE Journal · 1st AuthorAI Therapist realizing Expert Verbal Cues for Effective Robot-Assisted Gait Training
Minsu Chang, Tae-woo Kim, Jaewon Beom, Sunjae Won, Doyoung Jeon
IEEE Transactions on Neural Systems & Rehabilitation Engineering (TNSRE), Vol. 28, No. 12, Nov 2020 · DOI: 10.1109/TNSRE.2020.3038175
IEEE Xplore ↗
TNSRE 19
★ Frontier IEEE Journal · 1st AuthorResearch on Terrain Identification of the Smart Prosthetic Ankle by Fuzzy Logic
Minsu Chang, Kyungsoon Kim, Doyoung Jeon
IEEE Transactions on Neural Systems & Rehabilitation Engineering (TNSRE), Vol. 27, No. 9, Aug 2019 · DOI: 10.1109/TNSRE.2019.2933874
IEEE Xplore ↗
JKROS 26
★ Robotics Journal · 2026Learning-Based Sampling Strategies Leveraging Bidirectional Motion Planning Characteristics
Suhyun Jeon, M. Kim, Jun-Won Jang, Minsu Chang, J. Park
The Journal of Korea Robotics Society (JKROS), Vol. 21, No. 2, pp. 164–171, 2026
Google Scholar ↗
BioRob 20
★ Flagship IEEE Conf · 1st AuthorDevelopment of the AI therapist estimating the Borg scale of stroke patients in robot-assisted gait training
Minsu Chang, Doyoung Jeon
The 8th IEEE RAS/EMBS International Conference on Biomedical Robotics & Biomechatronics (BioRob), New York, 2020
Google Scholar ↗
ICORR 17
★ Flagship IEEE Conf · 1st AuthorA research on the postural stability of a person wearing the lower limb exoskeletal robot by the HAT model
Minsu Chang, Yeongmin Kim, Yoseph Lee, Doyoung Jeon
2017 IEEE International Conference on Rehabilitation Robotics (ICORR) in RehabWeek, pp. 369–374, London, 2017 · DOI: 10.1109/ICORR.2017.8009275
IEEE Xplore ↗
SCIE 19
★ SCIE JournalPrecise flowrate control of fluid gear pumps in automated painting systems using a repetitive controller
Kiyang Park, Minsu Chang, Doyoung Jeon
Applied Sciences, Vol. 9, No. 16, 3413, Aug 2019 · DOI: 10.3390/app9163413
MDPI ↗
arXiv 25
The Realization of Virtual Environments in the Lower Limb Exoskeletal Robot
Minsu Chang, Doyoung Jeon
arXiv preprint arXiv:2503.16459, 2025
arXiv ↗
ICCAS 19
Intelligent Impedance Control Using Reinforcement Learning for the effective Robot-Assisted Rehabilitation
Minsu Chang, Doyoung Jeon
The 19th International Conference on Control, Automation and Systems (ICCAS 2019), Jeju, 2019
ICCAS 19
AI Therapist realizing the Expert's Verbal Cue for the Effective Robotic Gait Rehabilitation
Minsu Chang, Doyoung Jeon
The 19th International Conference on Control, Automation and Systems (ICCAS 2019), Jeju, 2019
INRS 19
Development of the AI Therapist realizing the Expert’s Verbal Cue for the Effective Robotic Gait Rehabilitation
Minsu Chang, Doyoung Jeon
International Neurorehabilitation Symposium (INRS) in RehabWeek, Toronto, 2019
ICNR 18
The AI Supervisor for the effective treadmill training system of rehabilitation and exercise
Jaeyoung Kim, Minsu Chang, Doyoung Jeon
4th International Conference on NeuroRehabilitation (ICNR 2018), Pisa, 2018
ICCAS 18
An Expert System for the Effective Treadmill Training System of Rehabilitation and Exercise by using ANFIS
Minsu Chang, Doyoung Jeon
18th International Conference on Control, Automation and Systems (ICCAS), Pyeongchang, 2018
Thesis 18
능동보행훈련을 위한 하지외골격로봇의 필요에 따른 보조 제어 방법: An assist-as-needed control method of a lower limb exoskeletal robot for active gait training
장민수 (Minsu Chang)
서강대학교 대학원 (Sogang University Graduate School), 2018
Google Scholar ↗
ICCAS 17
The Effective Rehabilitation Method Realizing Various Environments in Robot-assisted Gait Training 🏆 (Best Presentation Award)
Minsu Chang, Doyoung Jeon
17th International Conference on Control, Automation and Systems (ICCAS), Jeju, 2017 · Best Presentation Award for Interactive Poster Sessions
ICCAS 17
A method of specific muscle group training of lower limb using an exoskeletal robot
Minsu Chang, Doyoung Jeon
17th International Conference on Control, Automation and Systems (ICCAS), Jeju, 2017
ICCAS 17
Realization of Virtual Impedance in the Lower Limb Assistive Robot SUBAR
Yeongmin Kim, Minsu Chang, Doyoung Jeon
17th International Conference on Control, Automation and Systems (ICCAS), Jeju, 2017
INRS 17
Realization of the Effects of Various Environments in Robot-assisted Gait Training
Minsu Chang, Doyoung Jeon
International Neurorehabilitation Symposium (INRS) in RehabWeek, London, 2017
KSPE 16
하지외골격로봇 착용자의 자세 안정성 분석을 위한 HAT Model 의 실험적 검증 (Experimental Verification of the HAT Model for Postural Stability Analysis of a Lower-Limb Exoskeleton Wearer)
장민수 (Minsu Chang), 전도영 (Doyoung Jeon)
한국정밀공학회 학술발표대회 논문집 (KSPE), pp. 83–84, 2016
Google Scholar ↗
KROS 16
보행재활로봇의 개발 경험과 이슈 (Development Experience and Issues of Gait Rehabilitation Robots)
장민수 (Minsu Chang), 전도영 (Doyoung Jeon)
로봇과 인간 (Robotics and Human), Vol. 13, No. 2, pp. 14–20, 2016
DBpia ↗
INTELLECTUAL PROPERTY · 10 US PATENTS/APPLICATIONS + 4 KR REGISTERED PATENTS

14 Patents — Autonomous Mobile Robotics, Lab Automation & Exoskeleton Control

Spanning autonomous mobile robot path planning, safety-margin adaptation, powder-dispensing manipulation, and material synthesis systems at Samsung Electronics (SAIT) to KFDA-certified lower-limb rehabilitation exoskeletons at Sogang University.

P01
Open-close type spatula, powder-dispensing apparatus including the same, and method of controlling powder-dispensing apparatus
Minsu CHANG, Jun-Won JANG, Hyun Do CHOI, Ki-Won GO, Won Je CHOI, Daewoong HAN
US Patent Application No. 19/233,597 (2026)
Google Scholar ↗
P02
Electronic device and method of planning moving path of robot
Jun-Won JANG, J. Park, Suhyun JEON, M. Kim, Minsu CHANG
US Patent Application No. 19/249,393 (2026)
Google Scholar ↗
P03
Mobile machine for adjusting safety margin and moving method thereof
Minsu CHANG, SeungYeon Kim, JANG Jun-Won, Won Je CHOI, Daewoong Han, PARK HyunKyu
US Patent Application No. 19/305,009 · Pub. No. US20260072439A1 (2026)
Google Patents ↗
P04
Transfer system
Minsu CHANG, SeungYeon Kim, JANG Jun-Won, Daewoong Han, Hyun Do CHOI, HA Taesin
US Patent Application No. 19/041,483 · Pub. No. US20260028212A1 (2026)
Google Patents ↗
P05
Separation and purification module and automated material reaction system including the same
Hyun Do CHOI, J. Kim, H. Kwon, B. Kim, J.W. Kim, D. LEE, S. Lee, Minsu CHANG, et al.
US Patent Application No. 19/068,604 (2025)
Google Scholar ↗
P06
Optical monitoring device
Wooram HONG, Hyukju Kwon, Gahee Kim, Won Seok Oh, Hyunjeong JEON, Hyun Do CHOI, Minsu CHANG, Younsuk Choi
US Granted Patent No. US12411092B2 (2025)
Google Patents ↗
P07
Material synthesis device, material synthesis platform, and method of operating material synthesis device
Sangyoon Lee, Dongseon LEE, Minsu CHANG, Jaejun Chang, Jaehyung Jang, Taesin HA
US Patent Application No. 18/989,808 · Pub. No. US20250232845A1 (2025)
Google Patents ↗
P08
Automatic evaporator system
Hyun Do CHOI, Bosung Kim, SeungYeon Kim, Seoungjin SEO, Minsu CHANG, Jun-Won JANG, Hyunjeong JEON, Wooram HONG
US Patent Application No. 18/968,676 · Pub. No. US20250180449A1 (2025)
Google Patents ↗
P09
Liquid processing apparatus and liquid processing system including the same
SeungYeon Kim, Bosung Kim, Seoungjin SEO, Minsu CHANG, Jun-Won JANG, Hyunjeong JEON, Hyun Do CHOI, Dai HEO, Wooram HONG
US Patent Application No. 18/885,072 · Pub. No. US20250180450A1 (2025)
Google Patents ↗
P10
Platform system
Daewoong Han, SeungYeon Kim, Minsu CHANG, Jun-Won JANG, Taesin HA
US Patent Application No. 18/908,258 · Pub. No. US20250122061A1 (2025)
Google Patents ↗
P11
Robot of assisting to exercise one's lower body, the production method of it, and the controlling method of it
JEON, Do Young, CHANG, Min-su
KR Registered Patent No. 10-2262191 (Pub. KR20200142384A)
Google Patents ↗
P12
Robot for assisting muscular strength in walking
JEON, Do Young, CHANG, Min-su
KR Registered Patent No. 10-2218260 (Pub. KR20200142347A)
Google Patents ↗
P13
REHABILITATION ROBOT FOR GAIT ASSISTANT DRIVEN BY A MOTOR
CHANG, Min-su, JEON, Do Young
KR Registered Patent No. 10-2132113 (KR102132113B1)
Google Patents ↗
P14
STABILIZATION CONTROL METHOD FOR EXOSKELETON ROBOT OF LEGS
SEO, Chang-Hoon, JEON, Do-Young, CHANG, Min-Su
KR Registered Patent No. 10-1940901 (KR101940901B1)
Google Patents ↗
RESEARCH (~2022) · KFDA MEDICAL EXOSKELETON (6 HOSPITALS) & AI THERAPIST

Gait Rehabilitation Exoskeletons (SUBAR) & Clinical AI Therapist

I earned my B.S. (2016), M.S. (2018), and Ph.D. (2022) degrees in Mechanical Engineering from Sogang University, Seoul, South Korea. My academic journey focused on hardware development and control algorithms for gait rehabilitation robots. Our research team achieved medical device certification from the Korea Food and Drug Administration (KFDA), enabling the deployment of our rehabilitation robots (SUBAR) in six hospitals across Korea. Over three years, we collected clinical robotic data from stroke patients and physical therapists, using machine learning and control algorithms to develop and implement an AI-driven therapist for gait rehabilitation.

Core Technical Stack (~2022): Control System Design, Machine Learning (ANFIS, PSO, RL), Human-Robot Interaction, Mechanical CAD (NX, SolidWorks), and real-time software in Python, C/C++, MATLAB/Simulink, LabVIEW, OpenSim, and ROS.

10 Clinical & Control System Video Demonstrations

1. AI Therapist (Ver. 1 — IEEE TNSRE 2020)
Implements the professional physical therapist's real-time verbal cues in robot-assisted gait training.
2. SUBAR Exoskeleton Robot (Hospital Deployment)
KFDA-certified lower-limb exoskeleton deployed across Korean hospitals for stroke patient gait rehabilitation.
3. SUBAR with AI Therapist (Ver. 2, Korean)
Realizes expert physical therapist decision-making and adaptive parameters during robot-assisted gait training.
4. How to Use SUBAR (Clinical UI Manual)
Walkthrough of SUBAR's clinical touchscreen user interface across operating modes and motions.
5. Lower-Limb Training with Visual Feedback (SMART 2D Game)
Interactive visual feedback system encouraging active patient participation during lower-limb training.
6. VR-Integrated Exoskeleton for Rehabilitation
Combines the lower-limb exoskeleton with an immersive Virtual Reality (VR) environment for engaging gait training.
7. Task-Oriented Training with Virtual Tunnel & Assist-as-Needed
EXOWheel visual feedback with virtual-tunnel impedance guidance assisting the patient only as needed.
8. AI Estimator for Stroke Patient Borg Scale (IEEE BioRob 2020)
Machine learning estimator predicting Rating of Perceived Exertion (Borg scale) from clinical and robotic sensor data.
9. Global Optimization (PSO & Neuro-Fuzzy ANFIS)
Particle Swarm Optimization (PSO) for training adaptive neuro-fuzzy inference systems.
10. Anthropometry-Based Control (OpenSim Musculoskeletal Simulation)
OpenSim biomechanical simulation used to design personalized exoskeleton control laws.

National TV Broadcasts & Korean Press Coverage

YonhapnewsTV — 고령화로 늘어나는 재활수요…의료기기 국산화 시급
National broadcast featuring our KFDA-certified gait rehabilitation exoskeleton in clinical use.
YTN Science — 중증 환자 재활 돕는다…'웨어러블 로봇'의 진화
Science broadcast highlighting wearable robot rehabilitation for severe stroke and burn patients.
📰 헬스조선: 중앙대병원, 지능형 하지재활로봇 치료 도입 ↗ 📰 중앙일보: 재활훈련 로봇이 도와드립니다 ↗ 📰 매일경제: 한림대 한강성심병원, 화상 로봇재활치료 국내 첫 시행 ↗ 📰 조선일보: 입는 로봇이 화상 환자의 재활 고통 줄인다 ↗

2021 Personal Project: NVIDIA Jetson Nano & JetBot Autonomous Driving

Self-directed embedded AI training project (Python, ROS, and Deep Learning on NVIDIA Jetson Nano) to reduce software development cost and bridge perception with mobile control:

Jetson Nano Project (1) — Assembling the Yahboom JetBot
Hardware build and board bring-up of the Jetson Nano mobile robot platform.
Jetson Nano Project (2) — Object Detection & Auto-Pilot First Trial
Real-time object detection and autonomous navigation using Python, ROS, and NVIDIA Jetson Nano.
TECHNICAL RECAP · GOOGLE CLOUD ONBOARD: GOOGLE SKILLS EDITION (WESTIN JOSUN PARNAS)

Architecting Enterprise-Grade Agentic Services on Vertex AI

What's the best way to follow up a solo win at the Google Gemini 3 Seoul Hackathon? By diving even deeper into the Google AI ecosystem! Building a rapid 7-hour prototype (GeminiSpace) is very different from building a scalable, enterprise-grade service. To take my multi-agent collaboration platform (Gouncil / M9G Agent Corps) to the next level, I attended the Google Cloud OnBoard — Google Skills Edition at the Westin Josun Parnas.

1. The Google AI Vision: Multimodality and Scale

Keynote sessions from Google Cloud and Google DeepMind leaders showcased the expansion of the Vertex AI Model Garden (200+ models including Gemini 3, Gemma 3, and open weights) alongside GenMedia on Vertex AI (Veo for video and Lyria for music). We are no longer just prompting text; we are orchestrating complex, multi-sensory data pipelines.

2. Track 1: AI Build-up — ADK, MCP vs. A2A Protocols & Cloud Run

Agent Development Kit (ADK) & Protocols: Learning the architectural distinction between the Model Context Protocol (MCP)—used to connect agents to tools and APIs—and the Agent2Agent (A2A) protocol for dynamic communication between distinct autonomous agents (e.g., a Travel Agent collaborating with a Flight Agent) provided the exact routing blueprint for the Gouncil platform.
Serverless AI on Cloud Run: Hosting LangGraph and ADK directly on Cloud Run, connecting to Vector Databases (AlloyDB / Cloud SQL), and automating CI/CD with Cloud Build.

3. The Highlight: 4-Layer Vertex AI Memory Architecture

Presented by Generative AI Field Solution Architect Sang-woon So, Vertex AI solves long-horizon context retention by separating agent memory into four human-like cognitive layers:

Memory LayerCognitive Role & Vertex AI Implementation
1. Session Memory (Short-Term)Captures contextual dialog state and tool return values during an active, ongoing session.
2. Personalized Memory Bank (Long-Term)Stores user preferences, profiles, and accumulated insights across multiple interactions over time.
3. External Information GroundingAccesses real-time world state and location-based facts via Google Search & Google Maps Grounding.
4. Internal Information RetrievalFetches enterprise/domain data using the RAG Engine and ScaNN-based Vertex AI Vector Search.
TEACHING ASSISTANT · SOGANG UNIVERSITY MECHANICAL ENGINEERING (2016 – 2020)

Control Theory & Robotics Teaching Assistant Lectures (LQR & MATLAB)

Nine consecutive semesters of undergraduate and graduate teaching assistantships at Sogang University, including online MATLAB & LQR control practice lectures recorded during the COVID-19 pandemic.

SemesterCourses Taught (Teaching Assistant)
2016-2Automatic Control
2017-1Automatic Control, Advanced Control I
2017-2Design and Control of the Robot, Experiments for Mechanical Engineering, Advanced Control II
2018-1Automatic Control, Advanced Control I
2018-2Design and Control of the Robot, Experiments for Mechanical Engineering
2019-1Automatic Control
2019-2Research Project I
2020-1Research Project II, Automatic Control
2020-2Advanced Control I
🎬 Full Online Control Lectures YouTube Playlist (PLqOt-5LGFKZuG2rmubGnM0TPNJxMo3oGx) ↗

7 Featured Online Control & MATLAB Practice Lectures (Korean)

1. Best of My Lectures — LQR Control (고등제어I LQR MATLAB 실습)
Linear Quadratic Regulator (LQR) state-feedback design and MATLAB simulation practice.
2. Advanced Control I / Robot Control (1/7) — 로봇설계제어 MATLAB 실습 1/3
Dynamic modeling and control simulation in MATLAB (Part 1).
3. Advanced Control I TA (2/7) — 로봇설계제어 MATLAB 실습 2/3
Manipulator trajectory tracking and feedback control in MATLAB (Part 2).
4. Research Project II (3/6) — 연구프로젝트II 3주차 2/2
Experimental system identification and control implementation guide.
5. Research Project II (5/6) — 연구프로젝트II 5주차 실습
Week 5 hands-on engineering lab session.
6. Automatic Control (1/4) — 4/16 자동제어 실습 MATLAB 2/2
Classical root-locus, Bode plot, and PID controller design in MATLAB.
7. Automatic Control (2/4) — 5월 21일 자동제어 실습
Frequency-domain analysis and closed-loop stability verification in MATLAB.
GLOBAL LEADERSHIP · AIESEC EXCHANGE TEACHER IN HUNGARY (2014)

AIESEC Exchange Teacher in Hungary — Winter & Summer 2014

I visited Hungary twice as an exchange teacher—experiences that remain some of my most challenging yet rewarding memories prior to starting my robotics research career, gained through my two-year involvement with AIESEC.

💡 Winter 2014: The Magellan Project

At Ferenc Földes Secondary School (Miskolc, Hungary), I led 10 classes where I shared insights into Korean culture, history, and my own academic journey. Beyond the presentations, I found deep fulfillment in discussing life goals with the students. Living in a multicultural environment, I spent my free time with a diverse group of friends from Brazil, Italy, Taiwan, and Hungary, which broadened my global perspective.

💡 Summer 2014: English Youth Camp

Later that year, I returned to Hungary to collaborate with educators from nine different countries. My focus was on making English accessible and engaging; I designed interactive activities to help students feel comfortable with the language while facilitating discussions on everything from daily life to pressing global issues.