Evrim Yazgin
Cosmos science journalist
A team led by researchers from Melbourne’s RMIT University has produced a tiny breakthrough device that “sees” and creates memories in a similar way to human vision. The research is published in the journal Advanced Functional Materials.
The device is a potential step toward applications such as self-driving cars and bionic eyes. It is only three nanometres thick – thousands of times thinner than a human hair – and requires no external parts to operate.
Yet it can capture, process and store visual information, just like the optical system of a human’s eyes and brain. And it does all of this with a tiny amount of data, making it a prime candidate for applications that require ultra-fast decision making.
Read more: Watch a 12-legged “myriapod” robot trek about in calm, curved lines
“Performing all of these functions on one small device had proven to be a big challenge until now,” says team leader Professor Sumeet Walia, from RMIT’s School of Engineering. “We’ve made real-time decision making a possibility with our invention, because it doesn’t need to process large amounts of irrelevant data and it’s not being slowed down by data transfer to separate processors.”
The new device can retain information for longer periods than previous devices and is more energy efficient – it doesn’t need frequent electrical signals to refresh its memory.
RMIT PhD researcher Aishani Mazumder says the development team used the analogue processing of the human brain, which can process information quickly and efficiently using minimal energy, as a blueprint.
Devices that attempt to mimic brain function are called “neuromorphic”.
“By contrast, digital processing is energy and carbon intensive, and inhibits rapid information gathering and processing,” Mazumder adds. “Neuromorphic vision systems are designed to use similar analogue processing to the human brain, which can greatly reduce the amount of energy needed to perform complex visual tasks compared with today’s technologies.”
Human eyes have a single retina that captures an entire image, which is then processed by the brain to identify objects, colours and other features.
“The human eye is exceptionally adept at responding to changes in the surrounding environment in a faster and much more efficient way than cameras and computers currently can,” Walia says.
The team mimicked the retina by using single-element image sensors that can capture, store and process visual information.
Read more: Robot chef makes salad after watching cooking video
They did this by doping (inserting atoms or ions) into ultra-thin wafers of indium oxide (In2O3) to detect and analyse UV light. This proof-of-concept experiment will be expanded to include a broader spectrum of light, including visible light and infrared.
Such technology could be used in autonomous operations in dangerous environments, self-driving vehicles, tracking space junk, food shelf-life assessment and advanced forensics.
It may one day also lead to the development of a visual prosthetic, or “bionic eye”, to restore sight or even give sight to those who have never been able to see before. Such a development hinges on the ability of researchers to produce viable technologies that can assess huge amounts of visual information in real time – not an easy task, but one that studies such as this may pave the way toward.
“The platform is a step towards smart visual aids as one of many application areas and this will rely on the width of the light spectrum we can capture,” Walia says in an email to Cosmos.