Thinking ahead

A UK research project is attempting to apply lessons learned from one of the most complex systems of all — the human brain — to improve the efficiency of systems such as air traffic control.

The issue, according to neurology specialist Seth Bullock of SouthamptonUniversity, who is directing the research, is one of space. ‘In the brain, the neurons act like components in an electrical circuit, and signals pass between them very fast,’ he said.

‘But there’s another system at work. It turns out that the neurons are suspended in a kind of chemical soup, and their signal causes changes in that soup which generates new chemicals. They diffuse away and have an effect on the other cells they come into contact with. It’s a slower effect, but it’s like a broadcasting function.’

Bullock believes that many other complex systems have a similar structure, and this has been neglected in previous studies. ‘When people study networks, they’ll make a chart of all the different nodes that operate in the network and draw lines to show how they’re all connected. You end up with a beautiful filigree, but you’re throwing away all the information on the physical space that separates the nodes. We think these spatial aspects can be very important to the modulation of the network.’

Bullock’s three-year project, called Spatially Embedded Complex Systems Engineering (SECSE), aims to discover whether other systems have a neuronal structure, and whether this can be used to manage them more efficiently. ‘Many industries rely on networks with fast signalling to keep them running. But there’s always a spatial aspect to these systems as well. Can we make use of that?’

The air traffic control network is a good example of such a system, said Bullock. ‘It brings together people, technology, and space. Fast communication, such as between controllers and planes, is important, but there’s an obvious spatial factor as well.’

As air travel costs are squeezed and concerns grow about the environmental impact of jet engines, control systems need to become more efficient to ensure that aircraft spend as little time in the air as possible. This will conserve fuel, minimise emissions and keep costs down. But Bullock is keen to stress that his work is still at a blue-sky stage.

‘The research is rather abstract at the moment — we’re trying to use our results from neural network research to make abstract models which we can then apply in new applications,’ said Bullock.

‘But we think that we can give systems like air traffic control a new vocabulary to describe their problems, so they can look at them from a different angle and find solutions in places they hadn’t previously considered.’