CORVALLIS, Ore. – Inspired by the human brain, Oregon State University researchers have developed a new light-sensitive device that combines sensing and memory while controlling how digital memories strengthen or fade over time.
Technology that functions more like the human brain could enable artificial intelligence systems to work faster while consuming less electricity, said project leader Larry Cheng of the OSU College of Engineering.
The new device integrates light sensing, memory and signal processing in a single phototransistor. Current AI hardware, Cheng explains, typically spreads those functions among different components, requiring information to move between them, increasing energy demands and reducing efficiency.
"Our optoelectronic device introduces a new hardware capability that may enable more efficient processing of information directly at the sensor level," said Cheng, professor of electrical engineering and computer science. "Unlike conventional memory that is designed to preserve information, our device can electronically control how memories strengthen or decay."
In the new device, light creates stored electrical charges that act as memory. Similar to how chemical signals in the brain regulate memory strength and forgetting, a small electrical signal adjusts the influence of those stored charges, allowing memories to persist longer or fade more quickly – an important steppingstone toward neuromorphic computing systems, Cheng said.
Scientists are exploring neuromorphic computing, which is modeled after the structure and function of biological neural systems, and in-sensor computing for their potential to process dynamic information more efficiently.
"This light-sensitive memory with a programmable memory lifetime creates a tunable time window for processing visual and other sensor signals directly where they are detected, a capability that could enable more efficient vision systems and other sensor-based AI technologies," Cheng said.
The device works by melding two different materials performing distinct functions. An oxide semiconductor serves as the transistor channel that carries electrical current, while an organic photosensitive material on top absorbs light and generates electrical charges.
Some of those charges become trapped within the photosensitive layer. These trapped charges continue to influence the current flowing through the oxide semiconductor even after the light is removed, allowing the device to retain a memory of past optical signals.
"What makes this work unique is that the stored charges are not fixed in place," Cheng said.
By applying an electrical gate voltage, the position of the trapped charges relative to the transistor channel can be changed. Moving the charges closer to the transistor channel – the microscopic pathway through which electrons flow – strengthens their electrical influence and prolongs the memory effect, he said. Moving them farther away weakens that influence and speeds the loss of stored charges, causing the memory to fade more quickly.
The National Science Foundation supported the research, which was published in Advanced Functional Materials . Collaborators included Ahasan Ullah, Tasnim Sarker, Xueqiao Zhang, Andrew Ensinger and Lizhong Chen of the OSU College of Engineering, and Roshell Lamug and Oksana Ostroverkhova of the OSU College of Science.
