Grundbrakkin Vision System Mimmicks Brain tae Cut E-Waste

A new energy-efficient artificial vision system, inspired by da human brain an pairtly made frae hinny, promises tae lessen da burden o electronic waste. Engineers frae da University o Glesga, in collaboration wi researchers frae São Paulo State University an Hong Kong Metropolitan University, hae crafted a system dat employs organic, biodegradable, an recyclable materials. Dis innovative design allous da system tae ‘see’ an ‘mind’ colours while consumin minimal pouer.

Da device, kent as an Electrolyte-Gated Organic Field-Effect Transistor (EGOFET), integrates three essential functions: sensin licht, processin information, an storin it in memory. Remerkably, it retains information even whan pouered aff, shawcasin a feature caad non-volatility. Dis capability cuid mak it pairticularly effective for real-warld uses, sic as in autonomous drones or smairt security systems.

Da EGOFET improves upon previous artificial vision systems dat relied on silicon-based CMOS sensors, dat demand substantial computin pouer an energy. Theodoros Serghiou, frae da University o Glesga’s James Watt School o Engineerin, led da development. He explained, “In conventional computin, dere’s an inherent latency frae haein tae fetch an transfer data in CMOS-based systems due tae da physical separation atween da processin an memory units. Oor new memory-based device, houaniver, performs dese functions simultaneously in-memory, similar tae hoo synapses in da human brain wark.”

Da team biggit dere prototype on a glass substrate uisin gowd electrodes, a layer o organic, photosensitive perylene, an hinny as an electrolyte. Dis setup allous da system tae act as a photodetector, producin current spikes dat vary wi different licht wavelengths an intensities. Da EGOFET staunds oot for its energy efficiency, requirin ainly 2.4 picojoules o energy per spikin event, makin it ane o da maist efficient devices o its kind.

Professor Jeff Kettle, da paper’s correspondin author, noted, “Oor device is able tae emulate key synaptic behaviours sic as short-term an lang-term plasticity wi heich fidelity.” Da research has implications ayont artificial vision, extendin intae sustainable neuromorphic computin an energy-efficient electronics. Da team plans tae scale da prototype intae arrays for better image recognition, pavin da wey for eco-friendly artificial vision systems.

Aince da device raxes da eyn o its uisefu life, da glass an gowd components can be recycled, while da organic pairts will biodegrade naiturally. Da research, cryed ‘Sustainable an Tunable Synaptic Electrolyte-Gated Organic Field-Effect Transistors (EGOFETs) for Licht Adaptive Visual Perceptive Systems,’ is published in Advanced Functional Materials. It received support frae da UKRI Responsible Electronics an Circular Technology Centre an da São Paulo Research Foundation in Brazil.