Researchers have tried creating several articifial intelligence (AI) models in the image of the human brain to carry out different functions. In a similar quest, the researchers and scientists of the Nanyang Technological University, Singapore (NTU Singapore) have developed an AI-based 'brain' for robots which gives them the ability to sense pain and 'heal' themselves when damaged. 

Co-lead author of the study, Associate Professor Arindam Basu from the School of Electrical & Electronic Engineering said, "For robots to work together with humans one day, one concern is how to ensure they will interact safely with us. For that reason, scientists around the world have been finding ways to bring a sense of awareness to robots, such as being able to 'feel' pain, to react to it, and to withstand harsh operating conditions. However, the complexity of putting together the multitude of sensors required and the resultant fragility of such a system is a major barrier for widespread adoption."

The 'brain' is a system that has different sensory nodes that are attached to an AI algorithm; these nodes 'react to pain' and process the 'pain' by calculating the force of impact or pressure. This power enables the robot to understand where exactly it has sustained damage and if the damage is slight, begin repairing itself without human intervention. 

For the healing purpose, the robots have a 'self-healing gel', an ion-gel material, which enables the robot to repair its mechanical functions without human intervention if it gets 'injured' from a sharp cut. 

As the system is in its nascent stage, the robots are heavily wired, with exposed wiring because they do not process the information within the device. The sensory nodes are spread through the robot's 'skin' which sends data to an external central processing unit through a network of sensors, where the information is processed via machine learning methods. However, as a novel appraoch, the scietists have also connected the nodes to multiple processing units, allowing the robots to learn locally. This approach can help make the robots more compact and increase their response time. 

As an experiment, a robot was fixed with memtransitors, an experimental multi-terminal electronic component that might be used in the construction of artificial neural networks. The memtransitors were trained to act as artificial pain receptors and synapses to help the robot get trained to recognise pain and learn damaging stimuli. 

As a result, the robot learned to respond to 'inury' or pressure in real-time as well as continued to respond to the 'injury' after the damage was done, ensuring that the system  

“Our work has demonstrated the feasibility of a robotic system that is capable of processing information efficiently with minimal wiring and circuits. By reducing the number of electronic components required, our system should become affordable and scalable. This will help accelerate the adoption of a new generation of robots in the marketplace,” said Assoc Prof Basu, who is a neuromorphic computing expert.

The discovery could open pathways to creating robots that can mimick human neurobiological functions.