Robots that currently exist on the market typically use sensors for navigating their environment and performing different tasks. The sensors, however, do not process the data themselves, but sent it to a single processing unit. This results in the need for heavy wiring, delayed response times, and costly – let alone lengthy – maintenance and repair.
To address these issues, researchers from the Nanyang Technological University in Singapore have recently developed a new, brain-inspired approach whereby a robot is rigged with a network of sensor nodes, connected to multiple processing units that function like “mini-brains” on the robotic skin.
This not only means that learning takes place locally, but also reduces the requirement for cumbersome wires, and decreases response times by five- to ten-fold. In addition, such robots could be equipped with self-healing ion gel that would allow them to perform necessary repairs without human intervention.
A study detailing the new system was published in the prestigious academic journal Nature Communications.
To enable robots to sense “pain” from mechanical pressure, the researchers developed memtransistors – electronic devices capable of memory and information processing – that function as artificial pain receptors and synapses.
A series of experiments have shown that once the robot is “injured”, the molecules in the ion gel begin to interact, thereby causing the robot to repair the damaged site and restore its function without losing overall responsiveness.
“In this work, our team has taken an approach that is off-the-beaten path, by applying new learning materials, devices and fabrication methods for robots to mimic the human neuro-biological functions. While still at a prototype stage, our findings have laid down important frameworks for the field, pointing the way forward for researchers to tackle these challenges,” said co-lead author Professor Nripan Mathews.
The team is now looking for governmental and industry partners to make the new system suitable for large-scale deployment.