Integrated M.S.-Ph.D. Student in Artificial Intelligence at Hanyang University
I build learning-based robotic systems that connect simulation, optimization, control, and real-world deployment.
I am an Integrated M.S.-Ph.D. student in Artificial Intelligence at Hanyang University, working on robotics, AI control, reinforcement learning, autonomous systems, and learning-based robot intelligence.
My main focus is building robotic systems that move beyond isolated algorithms: from simulation and optimization to ROS2-based deployment, edge AI systems, and real-world robot experiments.
Robotics AI · Control · VLA/VLM · ROS2 · UAVs · Sensor Optimization · Edge Systems
| Area | Focus |
|---|---|
| Robotics & Autonomous Systems | ROS2, robot software architecture, real-world deployment |
| AI Control & Reinforcement Learning | learning-based control, policy learning, adaptive systems |
| Vision-Language-Action Models | robot manipulation, imitation learning, VLA/VLM-based control |
| UAV Dynamics Learning | few-shot adaptation, residual learning, simulation-to-target transfer |
| Sensor Placement Optimization | LiDAR deployment, MILP, Greedy, MAPPO, infrastructure sensing |
| Edge AI Systems | Jetson, Android interfaces, real-time monitoring, Wi-Fi streaming |
Optimization-based LiDAR placement research using MILP, Greedy, and learning-based methods for autonomous infrastructure sensing.
- Multi-LiDAR placement optimization
- Coverage-based deployment evaluation
- Comparison between mathematical optimization and learning-based approaches
Real-time Jetson-Android monitoring system for mobility platforms.
- ROS2-based status monitoring
- Android UI for edge device visualization
- Wi-Fi communication and real-time video streaming pipeline
- Jetson-based edge AI integration
Computer-vision learning examples covering YOLO detection, segmentation, Google Colab webcam inference, YouTube video inference, and VLM analysis.
- KITTI-format label conversion and YOLO training workflow
- Real-time Colab webcam inference examples
- VLM-based image understanding workflow
Drone response prediction framework using source selection, residual learning, and few-shot target adaptation.
- PX4 / MATLAB simulation data
- Reduced-order UAV dynamics modeling
- Source selection for transfer learning
- Residual adaptation with limited target data
Robot manipulation experiments using uFactory Lite6, Isaac Sim, and real-world demonstrations.
- Vision-Language-Action model experiments
- Real robot data collection
- Pick-and-place task learning
- Simulation-to-real manipulation pipeline
Public repositories and reproducible artifacts for these work-in-progress projects are to be added.
> Learning-based UAV response prediction
> VLA-based robot arm manipulation
> ROS2-based autonomous mobility systems
> Multi-sensor placement optimization
> Jetson-based real-time robot monitoring
| Project Type | Keywords |
|---|---|
| Robotics Software | ROS2, Jetson, Android, real-time monitoring |
| Autonomous Systems | LiDAR placement, UAV dynamics, sensor optimization |
| Learning Systems | RL, VLA/VLM, residual learning, imitation learning |
| Simulation | Isaac Sim, PX4, MATLAB, vehicle simulation |
- GitHub: @Geppetto0608
- Email: phj990608@hanyang.ac.kr