Action Selection and Individuation in Agent Based Modelling

This paper is a tutorial on action selection for Agent-Based Modelling (ABM). Having a clear idea of how you are organizing your agent’s intelligence will make your code cleaner and easier to maintain, and your models easier to communicate to others. This paper describes four means of orgainizing agent action selection in increasing order of complexity These are: environmental determinism, finite state machines, basic reactive plans, and Parallel-rooted, Ordered, Slipstack Hierarchical (POSH) reactive plans. Modellers should use the simplest mechanism possible — this paper describes the contexts in which more complicated mechanisms may be required, as well as suggesting coding and commenting schemes for all four systems. This paper also addresses the issue of Individuated Agent-Based Modelling (IABM), where individual agents display different behavior. It gives examples of existing IABM systems and describes how these can be moved into more mainstream ABM simulators via two relatively simple mechanisms: either exploiting individual local variables or by specifying different priorities within the action selection mechanism. This allows individual agents to vary in their behavior while sharing the vast majority of their code.