The tiny device mimics human vision and memory abilities
(Nanowerk News) Researchers have created tiny devices that ‘see’ and create memories in a similar way to humans, in a move that promises to one day have applications that can make quick, complex decisions like in self-driving cars.
The neuromorphic invention is a single chip activated by a sensing element, treated indium oxide, which is thousands of times thinner than a human hair and requires no external parts to operate.
RMIT University engineers in Australia led the work, with contributions from researchers at Deakin University and the University of Melbourne.
The team’s research demonstrates work tools that capture, process and store visual information. With precise engineering of doped indium oxide, these devices mimic the ability of the human eye to perceive light, package and transmit information like the optic nerve, and store and classify it in memory systems like our brains can.
Collectively, these functions can enable very fast decision making, the team said.
Team leader Professor Sumeet Walia said the new device could perform all the necessary functions – sensing, creating and processing information, and storing memories – rather than relying on external energy-intensive computing, which prevents real-time decision making.
“Performing all these functions on one small device has proven to be a huge challenge to date,” said Walia of RMIT’s School of Engineering.
“We have made real-time decision-making possible with our invention, as it does not need to process large amounts of irrelevant data and is not slowed down by transferring data to a separate processor.”
What did the team achieve and how did the technology work?
The new device is able to demonstrate the ability to retain information for a longer period of time, compared to previously reported devices, without the need for frequent electrical signals to refresh the memory. This capability significantly reduces energy consumption and improves device performance.
Their findings and analysis are published in Advanced Functional Materials (“Long duration persistent photocurrent in 3 nm thin doped indium oxide for integrated light sensing and in-sensor neuromorphic computation”).
First author and RMIT PhD researcher Aishani Mazumder said the human brain uses analog processing, which allows it to process information quickly and efficiently using minimal energy.
“In contrast, digital processing is energy and carbon intensive, and hinders the rapid collection and processing of information,” he said.
“The neuromorphic vision system is designed to use analog processing similar to that of the human brain, which can greatly reduce the amount of energy required to perform complex visual tasks compared to current technology.
What are the potential applications?
The team used ultraviolet light as part of their experiment, and are working to extend this technology even further to visible and infrared light – with many possible applications such as bionic vision, autonomous operation in hazardous environments, shelf life assessment of food and advanced forensics.
“Imagine a self-driving car that can see and recognize objects on the road in the same way a human driver can or can detect and quickly track space junk. This will be possible with neuromorphic vision technology.”
Walia said the neuromorphic system can adapt to new situations over time, becoming more efficient with more experience.
“Traditional computer vision systems – which cannot be miniaturized like neuromorphic technologies – are usually programmed with specific rules and don’t adapt easily,” he says.
“Neuromorphic robots have the potential to walk autonomously for long periods of time, in hazardous situations where workers are exposed to possible collapse, explosions and toxic air.”
The human eye has a single retina that captures the entire image, which is then processed by the brain to identify objects, colors and other visual features.
The team’s device mimics the capabilities of the retina by using a single element image sensor that captures, stores and processes visual information on a single platform, said Walia.
“The human eye is very adept at responding to changes in the surrounding environment in a way that is faster and far more efficient than today’s cameras and computers,” he said.
“Taking inspiration from the eye, we have worked for several years to create a camera with similar capabilities, through a process of neuromorphic engineering.”
