DE-TW-NeuroMemSense
Memristive Edge AI Chips for Trustworthy and Secure Biosensing
DE-TW-NeuroMemSense develops memristive edge AI chips for trustworthy and secure biosensing. The combination of neuromorphic architectures with resistive memory technologies opens new possibilities for intelligent, low-power sensor systems.
Objectives and Approach
The project aims to efficiently and securely monitor health data and process it in real time using AI methods. The innovative co-design approach combines fast analog hardware accelerators based on novel memristors with digital logic for algorithm training. Validation uses a sensor chip for precise blood pressure measurements and a 2D image sensor for pulse detection.
Innovations and Perspectives
The project deepens German-Taiwanese collaboration in microelectronics, strengthens the scientific network, and supports skilled workforce development in chip design. By delivering a memristive AI sensor chip for biomedical applications, it enhances innovation capacity at German research institutions.
Project Coordinator
- Universität Siegen
Project Partners
- National Yang Ming Chiao Tung University, Hsinchu
This project is funded by the German Federal Ministry for Research, Technology and Space (BMFTR) as part of the Design Initiative Microelectronics.
