Scientists from Tokyo University of Science have developed a groundbreaking device that mimics human synaptic behavior for efficient edge AI processing. The self-powered dye-sensitized solar cell-based optoelectronic device features controllable time constants, allowing it to process time-series data across multiple timescales. The device integrates optical input, AI computation, analog output, and power supply functions at the material level.
It exhibits synaptic characteristics such as paired-pulse facilitation and paired-pulse depression, which are essential for high computational performance in time-series data processing tasks. Associate Professor Takashi Ikuno, who led the research team, explained that the afterimage phenomenon of the human eye inspired the design of this optoelectronic synaptic device. The device has wide-ranging practical applications, including surveillance cameras, car cameras, and health monitoring devices.
When used as the reservoir layer of physical reservoir computing (PRC), the device demonstrated high accuracy in classifying human movements like bending, jumping, running, and walking, maintaining over 90% accuracy rates. Remarkably, the power consumption is just 1% of what conventional systems require, substantially lowering associated carbon emissions.
Novel device mimics synaptic behavior
Dr. Ikuno highlighted the device’s potential to be a popular edge AI optical sensor that can be attached to various objects or individuals, contributing to significant power consumption improvements in vehicles and other technologies. The findings were published in the journal ACS Applied Materials & Interfaces on October 28, 2024.
The research was partly funded by the Japan Science and Technology Agency (JST) and benefited from technical support with IPCE measurements from Mr. Tatsuya Yamamoto, Mr. Naoki Kiyota, and Prof.
Morio Nagata of Tokyo University of Science. This novel device presents significant advancements in energy-efficient edge AI sensors, promising varied applications while reducing power consumption costs. It underscores a major step forward in the integration of artificial intelligence with sustainable practices.
Cameron is a highly regarded contributor in the rapidly evolving fields of artificial intelligence (AI) and machine learning. His articles delve into the theoretical underpinnings of AI, the practical applications of machine learning across industries, ethical considerations of autonomous systems, and the societal impacts of these disruptive technologies.
