Agent based modelling (ABM) uses autonomous 'agents' as its functioning components. These agents act as unique individuals, each with their own 'knowledge' or memory and each with a set of rules governing their roles and interactions. Different agents can have different roles. Each agent can have its own private information and/or public information which it shares with all or some of its neighbours. Agents react to each other and their surrounding 'environment'. They can be given instructions to modify their own internal states, other agents states and/or their environment, thus directly or indirectly communicating. The environment may comprise of other agents or other types of object or data. For example, molecules and their specified concentrations could be environmental data in a microbiology simulation, or weather such as rain or wind in a model containing flocking 'birds' as agents.