EMERGING MEMORY DEVICES

   Our research aims to realize an energy-efficient in-sensor computing system by integrationg a highly sensitive TFT based gas sensor array with a non-volatile RRAM synaptic array. This co-integrated architecture enables real-time gas detection and on-sensor signal processing, allowing key features to be extracted directly at the sensor node. By minimizing data movement between sensing and computing blocks, the platform reduces power consumption and latency for edge operation. Ultinately, we target a scalable sensor-memory array platform that supports low-power, always-on monitoring and future neuromorphic/learning-capable sensing applications.