Brainwave

A robot controlled by a computer model that exactly replicates brain neurone structures could help researchers to find treatments for diseases such as Parkinson’s, Huntington’s, schizophrenia and Tourette’s syndrome.

The specially-developed neural network, which recreates both the healthy and damaged ‘circuits’ of a portion of the brain, is helping researchers at Sheffield University towards a better understanding of brain disease.

The computer model has been developed using years of neurological research data. It allows the robot, which has wheels and grippers, to be programmed to mimic the movement problems experienced by Parkinson’s and other brain disease sufferers.

Although much brain research has been conducted, until now it has been very difficult to test the data collated, said Dr Tony Prescott, one of a team of eight working on the network and robot at the university’s psychology department.

Researchers may theorise about brain function from the data collected but without models and motive systems to test them, the theories will mean very little, he said.’We have just started modelling sensor structures to feed into our model of the basal ganglia, which controls movement. We need these models to test the data collated over the years from magnetic resonance imaging scans of people’s brains, animal studies and the observations of diseased patients.’

The technology, also known as computational neuroscience, differs from existing artificial intelligence software, which is based on just the principles of brain operation. Prescott’s network exactly recreates the data processing neurones and the structures within the brain that link them.

The project team, known as the Adaptive Behaviour Research Group, has received £590,000 from the EPSRC to model the sensory input of rat brains, which are similar to the human brain.

The technology is being used for a project called ‘Robot control using a model of central structures in the vertebrate brain’. This aims to develop ‘biomimetic robot control’ based on the basal ganglia and the reticular formation – the parts of the brain involved in movement.

Researchers are also involved in a project known as ‘Whiskerbot’, a robot whisker-like sensor system, again based on the rat.

This was launched after the team noticed that the progress of some mobile robots was hampered by the lack of equipment for close-quarter sensing in enclosed spaces such as ducting or piping systems, underground structures, and the interior of disaster sites.

In addition to object detection, the whiskers could also help the robot to determine surface texture.

The project is being undertaken in collaboration with the Intelligent Autonomous Systems Laboratory at the University of the West of England, based in Bristol.