Artificial nerve offers boost for robot designs
US-Korean research creates touch-sensitive system that can recognise Braille. Joel F. Hooper reports.
Bridging the gap between biological systems and machines has come one step closer, with a report in the journal Science detailing a touch-sensitive artificial nerve. Capable of distinguishing Braille characters and even interfacing with a cockroach leg, the device could have wide-reaching applications in robotics and prosthetic limbs.
The artificial nerve, developed by scientists at Stanford University in the US and Seoul National Universities in Korea, brings together three components to mimic at action of a sensory nerve cell.
An organic polymer-based pressure sensor feeds information to a flexible ring oscillator, which functions as an electronic neuron. The signal generated is then fed to an artificial synaptic transistor, which integrates the outputs from multiple pressure sensors, modelled after the synapse between biological nerve cells. This synaptic transistor can then feed output to a computer, or can be directly linked to a biological nerve cell.
The polymer-based pressure sensors, which have been previously reported by co-author Zhenan Bao, professor of chemical engineering at Stanford, can detect the weight of a single flower petal weighing just 0.8 milligrams. This means that artificial skin imbedded with these cells could be even more sensitive than human skin.
“We take skin for granted but it's a complex sensing, signalling and decision-making system,” says Bao.
“This artificial sensory nerve system is a step toward making skin-like sensory neural networks for all sorts of applications.”
While the goal of creating artificial skin is still a long way off, the team was able to demonstrate its artificial nerve in both sensing and reflex-like applications.
Six pressure sensors were able to distinguish different Braille letters that were pressed against them. When the signals from the sensors were integrated through the synapse-like transistor, the characters became more easily distinguished than when the signals were not integrated.
This mimics the role of tactile information processing performed by human nerve cells, where the input from multiple signals is integrated and partially processed before being delivered to the brain.
“Biological synapses can relay signals, and also store information to make simple decisions,” says co-author Tae-Woo Lee of Seoul National University.
“The synaptic transistor performs these functions in the artificial nerve circuit.”
In another experiment, the artificial nerve cell was directly connected to the nerve of a cockroach leg. As increasing pressure was applied to the sensor, the leg showed an increasing twitch response.
The authors liken this response to the reflexive jerk when our knee is tapped.
Reflex responses allow biological systems to react directly to stimuli, without having to wait for the brain the process the signal. A similar response in artificial systems could lead to rapid response times in robotics and artificial limbs and could help reduce power consumption by avoiding complex information processing.