Aloe Vera-Inspired Cognitive Computing: Unveiling the Power of Pavlovian Conditioning and Pattern Recognition with a Synaptic RRAM Device

Abstract

Biomaterial-based devices have demonstrated versatility in various applications and have drawn considerable attention. In the present work, we report on analog bipolar resistive switching characteristics of the naturally extracted aloe vera-based RRAM device for neuromorphic application. Endurance and retentivity of the Ag/aloe vera/FTO-based Bio-RRAM device demonstrated good stability and nonvolatile nature. Further synaptic learning rules such as long-term potentiation, long-term depression, and spike-time-dependent plasticity are demonstrated. The device presented good potentiation/depression stability with a low cycle-to-cycle variation of 4%. The Pavlovian conditioning on the Ag/aloe vera/FTO device is also demonstrated. These results suggest that the Ag/aloe vera/FTO synaptic Bio-RRAM device would indeed be a potential candidate for neuromorphic application. We used this device for a pattern recognition task, which is used to identify house numbers from the Google Street View house number (SVHN) data set. The test accuracy remains above 70% until an image resolution of 16 x 16 is retained. This indicates that the synaptic device can be efficiently utilized in deep learning applications with reduced input data dimensions.