It is currently testing its system with widely available ‘e-puck’ robots — but believes the work could have applications for disaster and environmental monitoring, security and even planetary exploration.
Project lead Prof Yaochu Jin, chair in computational intelligence at Surrey, explained that the inspiration behind the project came from studying how organisms develop.
‘Around two-and-a-half years ago I attended a conference on systems biology that tries to understand how cells interact with each other in the process of morphogenesis — so starting with one fertilised single cell, which divides, then migrates and forms a final body plan.
’It’s a self-organising process with no centralised control mechanism. In robotics and other areas of engineering this is very desirable because systems with central control are quite fragile because if the central controller is defective then the whole system will fail.’
The team set to designing computational models to mimic this behaviour and found they could create groups of simulated robots that were able to adapt to their environment, reconfigure and self-repair.
They have now applied a simplified version of the system to a fleet of e-puck robots that were able to encircle and entrap two moving targets. Jin hopes to apply the system to various applications and said the most obvious area may be in satellites — which are increasingly being deployed in constellations.
‘Say we had some natural disaster and you want to focus on one particular area, you would need to reconfigure the satellites. I want a system that can decide by itself where each satellite should go to achieve the best accuracy,’ Jin said.
UK productivity hindered by digital skills deficit – report
This is a bit of a nebulous subject. There are several sub-disciplines of 'digital skills' which all need different approaches. ...