Australian scientists are one step closer to creating a bionic brain after developing a nano memory which is capable of storing information in the same way as their biological counterpart.
Researchers from Melbourne’s RMIT said the device, which is 10,000 times thinner than a human hair, imitates the brain’s complex engineering with its twisted mass of neural pathways.
Lead researcher Hussein Nili said creating the nano memory cell was the first step in the process of building artificial neuron networks.
“The development of these nano memory cells is a prerequisite for building these artificial neuron networks that are capable of matching the performance and functionality of their biological counterparts,” he said.
While conventional digital storage, such as a USB, records data in a binary sequence of zeroes and ones, the nano memory cell can store information in multiple states because it is analogue.
Dr Nili likens the difference to that of a regular light switch and one that is fitted with a dimmer control.
“When you have a light switch at your home, you can just flip it on and off, so it’s either light or not light,” he said.
“But having a dimmer switch actually gives you much more flexibility in terms of how much light you’re using.
“So the analogue memory is kind of like that, compared to the digital memory.
“It can give you much more flexibility in terms of what information you store and what functionality you get.”
Dr Nili said it could lead to the development of a bionic brain which would remove the ethical barriers involved in experimenting on humans.
“If you can replicate a brain structure outside a human body, you can get more insights into the functionality of a mammalian or human brain, in terms of the kind of disorders that human brains develop, like Parkinson’s or Alzheimer’s disease,” he said.
While a long way off, the technology may one day be used to create replacement parts for people who suffer brain damage.
In the meantime, study co-author Sumeet Walia believes the cell could be used to develop artificial intelligence.
“Once we are capable of storing and remembering and recalling the past events, from there we can actually start working on developing it as a storage component for [the] full area of artificial intelligence networks,” Dr Walia said.
“For example, roberts, or even computers that behave like a human brain.”
The study has been published in the Advanced Functional Materials journal.