Ants and Agents: a Process Algebra Approach to Modelling Ant Colony Behaviour

Process algebras are widely used in the analysis of distributed computer systems. They allow formal reasoning about how the various components of a system contribute to its overall behaviour. In this paper we show how process algebras can be usefully applied to understanding social insect biology, in particular to studying the relationship between algorithmic behaviour of individual insects and the dynamical behaviour of their colony. We argue that process algebras provide a useful formalism for understanding this relationship, since they combine computer simulation, Markov chain analysis and mean-field methods of analysis. Indeed, process algebras can provide a framework for relating these three methods of analysis to each other and to experiments. We illustrate our approach with a series of graded examples of modelling activity in ant colonies.