Nanowire networks mimic the networked structure of the human brain. But can they learn and remember like a human brain can? New research indicates they can.
The brain’s powerful ability to process information is largely attributable to the network of connections formed by neurons and synapses. While we understand much of the brain's workings, some aspects, such as higher cognitive functions like learning and memory, remain elusive.
A kind of nanotechnology, nanowire networks (NWNs) are typically made from highly conductive silver wires, invisible to the naked eye, covered with plastic material and formed into a mesh. The nanowires self-assemble to form a dynamic and complex network that integrates memory and processing, similar to that seen in the brain.
Now, an international team, led by researchers from the University of Sydney, has proven just how similar NWNs are to the human brain.
“This nanowire network is like a synthetic neural network because the nanowires act like neurons, and the places where they connect with each other are analogous to synapses,” said Zdenka Kuncic, a co-author of the study.
To find out to what degree NWNs demonstrate cognitive functioning, researchers administered a version of a test that’s used to assess human working memory, called the n-back test.
Humans undertaking the n-back test might be shown a series of letters or images presented in a sequence. For each item in the sequence, they have to determine whether it matches the item presented ‘n’ items ago. An n-back score of seven is average, indicating that a person can recognize the item that appeared seven items ago.
For the NWN, the researchers modified the n-back test into implementable subtasks. To administer the test, the researchers guided the NWN’s pathways where they wanted.
“What we did here is manipulate the voltages of the end electrodes to force the pathways to change, rather than letting the network just do its own thing,” said Alon Loeffler, lead author of the study. “We forced the pathways to go where we wanted them to go.”
The researchers found that directing the NWN’s pathways improved its memory capacity and accuracy.
“When we implement that, its memory had much higher accuracy and didn’t really decrease over time, suggesting that we’ve found a way to strengthen the pathways to push them towards where we want them, and then the network remembers it,” Loeffler said.
The proof was in the testing. When they administered the modified n-back test to the NWN, it could ‘remember’ a desired endpoint in an electric circuit seven steps back, on par with human memory.
After constantly reinforcing the NWN, the researchers say, it reaches a point where the memory becomes fixed and no further reinforcement is needed.
“It’s kind of like the difference between long-term memory and short-term memory in our brains,” Kuncic said. “If we want to remember something for a long period of time, we really need to keep training our brains to consolidate that, otherwise it just kind of fades away over time.”
The researchers say that their study demonstrates that NWNs can operate in a similar way to the human brain and could be used to improve robotics or sensor devices that need to make decisions quickly.
“In this research, we found higher-order cognitive function, which we normally associate with the human brain, can be emulated in non-biological hardware,” Loeffler said. “Our current work paves the way towards replicating brain-like learning and memory in non-biological hardware systems and suggests that the underlying nature of brain-like intelligence may be physical.”
The study was published in the journal Science Advances.
Source: University of Sydney
