01 · COMPUTE
Computational Intelligence
Methods that learn, adapt, and optimize under practical limits.
- Efficient and edge machine learning
- Adaptive optimization
- Data-efficient learning
- Interpretable decision systems
Research program
Intelligence engineered for the real world.
We study systems where algorithms, signals, electronics, and physical action must work as one dependable whole.
02 · CONNECT
Embedded & Electrical Systems
Robust interfaces between physical signals and computational models.
- Embedded intelligence
- Smart sensing and signal processing
- Hardware–software co-design
- Energy-aware computing
03 · ACT
Autonomous Systems
Systems that perceive uncertainty and act safely in dynamic environments.
- Autonomous decision-making
- Robust and adaptive control
- Human–AI collaboration
- Real-world evaluation
04 · TRANSLATE
Applied Research
Research becomes valuable when it survives contact with real constraints.
- Prototype-driven inquiry
- Reproducible evaluation
- Cross-disciplinary collaboration
- Deployment-oriented design
Research approach
From question to working system
Our research process links four stages: formulate a measurable question, build a minimal testable system, evaluate under realistic constraints, and communicate the result in a reproducible form. We welcome collaboration around problems that genuinely require computation and electrical systems to be designed together.