Hyunjoong Lee (이현중)
email : ancelmo0403@gmail.com, guswnd0403@knu.ac.kr, guswnd0403@naver.com
CV (PDF)
email : ancelmo0403@gmail.com, guswnd0403@knu.ac.kr, guswnd0403@naver.com
CV (PDF)
About
I am an M.S. candidate (expected Feb 2027) in the School of Electronic and Electrical Engineering at Kyungpook National University (KNU), advised by Prof. Daejin Park in the AI-Embedded Software on Chip Lab.
I work on real-time embedded systems and timing-predictable automotive software for Software-Defined Vehicles. I take control and AI algorithms past the model stage onto automotive MCUs (Infineon AURIX TC375/TC397, NXP S32K344), real motors, and in-vehicle networks, and measure how they actually behave.
Education
Kyungpook National University, Daegu, Korea
M.S. in Electronic and Electrical Engineering (Mar 2025 – Feb 2027, expected). Advisor: Prof. Daejin Park.
Thesis in progress: On the Hardware–Software Gap in Runtime Control
B.S. in Electronics Engineering (Mar 2019 – Feb 2025)
Interests
Real-Time Embedded Systems
Timing-Predictable Automotive Software
Software-Defined Vehicles
ECU-in-the-Loop / Hardware-in-the-Loop Validation
Mixed-Criticality Control
HW/SW Co-Design
Embedded AI
Projects
Real-Time HILS for CNN-Based Motor Fault Diagnosis (Hyundai Motor Co. industry–academia project, 2026): Built an end-to-end HILS in which CARLA vehicle speed drives a real PMSM (NXP S32K344), a teammates' CNN on an AURIX TC397 diagnostic ECU classifies motor signals, and the result feeds back into motor control and CARLA. Built proxy signals from motor-control internals for a rig without the training sensors, implemented Goertzel-based feature extraction on the TC397, reproduced all five diagnosis classes through fault injection, and added a link-loss fail-safe and Hyundai-requested fault reactions (slow down / stop). Live-demonstrated at the Namyang R&D Center.
M.S. Thesis, in progress (2026–): On the Hardware–Software Gap in Runtime Control. A study of the instant a control result becomes fixed (binding time) when AI inference runs beside a real-time control loop, on a Xilinx Zynq-7020 (CPU + FPGA) and an NXP S32K344 driving a real motor.
Papers
Command-Authority Scheduling for Bounded Low-Criticality Service Influence in Mixed-Criticality SDV Control (IEEE Access, 2026)
JSEC: Jitter-to-Step Execution Contracts for Reproducible Distributed ECU-in-the-Loop Co-Simulation (IEEE TCAD, EMSOFT 2026 journal track, to appear)
Early Detection of Software-Induced Internal Margin Degradation in Automotive MCUs Using On-Chip ADC Variability (IPFA 2026)
Architecture-Aware Neural Compression for TriCore Firmware using Knowledge Distillation (ICAIIC 2026)
Real-Time Depth Map Using Stereo Vision and Convolutional Filtering (JKIICE, 2025)
Lightweight Empirical Reinforcement Learning Driven Adaptive Super-Twisting Control with Fused Linear-Nonlinear Sliding Surfaces for Embedded Vehicle Control (IECON 2025)
Advanced Real-Time Performance QEMU Emulated Automotive Control System with a Multithreaded PID Controller (ICCE-TW 2025)
Convolution-Based Depth Map With Shadow Removal Using Cameras for 3D Mapping in Autonomous Vehicle Driving (ISPACS 2024)
Auto Parking Assistance Control using Human-Activity Feedback-based Embedded Software Emulation (JKIICE, 2024)
Multi-Camera Interoperable Emulation Framework using Embedded Edge-Cloud AI Computing for Autonomous Vehicle Driving (VNC 2024)