All research areas
Sensors, Edge AI & IoT Systems
Resource-efficient sensing, embedded intelligence, and hardware/software co-design for always-on IoT and Edge-AI applications.
Our sensor and Edge-AI research focuses on systems that can acquire, interpret, and act on data close to where it is generated. The goal is to reduce communication and energy costs while enabling reliable, long-term operation on resource-constrained hardware.
Research directions
- Sensor interfaces and complete sensing systems
- Low-power Edge-AI and IoT hardware and software
- FPGA implementations of compact machine-learning models
- Hardware-aware recurrent neural networks for time-series data
- Embedded monitoring and classification under limited memory, power, and data
Previous work has explored low-power livestock-behavior monitoring with inertial sensors, data augmentation, and compact learning models. Current projects extend this hardware-aware approach to real-time FPGA implementations of RNN, GRU, and LSTM models for estimating the remaining useful life of lithium-ion batteries.