Electrospinning Device for Nanofibers| Frame-Channel Engineered Multimodal Transistor IntegratingNeuromorphic Computing with Cross-Modal Edge Intelligence

Challenge: The ongoing development of neuromorphic electronics aims to design electronic systems that mimic biological neural networks, addressing the hardware computing demands of supercomputers and the development of multimodal edge devices in the context of artificial intelligence. However, a key challenge in current neuromorphic devices replicating synaptic function lies in their over-reliance on non-volatile conductance updates to simulate neuromorphic behavior, which limits higher-order biomimetic capabilities and hinders modal expansion.

Approach: Professors Gang He & Shouguo Wang from Anhui University adopted a channel design using a double-layer nanofiber membrane (NM) composed of insulating and active materials, and innovatively constructed a field-effect transistor (FET) with a frame-channel (FC) structure, thereby establishing a more adaptable bio-inspired system.

Innovation 1: The FC architecture field-effect transistor (FCFET) significantly enhances the full-dimensional spontaneous coupling capability between active materials and physical spatial signals, activates excellent specific device characteristics, and demonstrates its universality.

Innovation 2: Combining light-stimulus-induced synaptic plasticity, a biologically inspired inhibitory integration regulation mechanism was achieved at the artificial synaptic device level, successfully demonstrating advanced biomimetic functions including synaptic regulation and lateral inhibition-based visual processing. Importantly, the device can cross-modally switch to an intelligent respiratory signal sensor and integrate with lightweight algorithms to develop efficient interactive edge systems.

https://doi.org/10.1002/adfm.202506332