Collective Animal Behaviour

graphic

We use honey bees, locusts and Mormon crickets as contrasting model systems to investigate the transfer of directional information within animal groups on the move. Our aim is to use this knowledge to produce a unified model for collective group movement. Such a model can be used to predict the movement of insect swarms, and applied more generally to understanding other collective phenomena such as human crowds, traffic control, and disease transmission epidemiology.

We further study problem solving in Nature, and by ants, bees and slime moulds in particular. Modern algorithmic techniques for solving optimization problems, such as ant colony optimization, genetic algorithms and particle swarm optimization, take their initial inspiration from biology. In some cases, mathematical models used in understanding the natural phenomena have aided design of algorithms, but in others biological inspiration is only superficial. As new fields such as autonomous and organic computing emerge, there is increasing demand for biologically inspired methods that provide robust and general solutions. We believe that the natural systems we study have exactly the properties required of autonomous computing systems: they can efficiently find optimal solutions to multiple problems in dynamic environments. But to investigate our claim we need to learn more about what specific dynamic problems are solved by social insects and slime moulds, and the mechanisms that allow them to produce these solutions.