2.2: Main elements of an ABM#
What is every agent-based model built from? However large or elaborate a model becomes, it is assembled from the same five kinds of ingredient: the agents that decide and act, the environment they live in, the passage of time that lets them act in turn, the interactions that link them, and the model that binds these together into one running whole. Knowing these five categories gives you a checklist you can bring to any model you meet or design, so that nothing important is left out.
Those five ingredients are the five sub-sections that follow:
The autonomous decision-makers of the model: their characteristics, roles, and state.
The world agents inhabit, and the grid, network, and continuous spaces they move through.
How simulations advance, and the different ways of scheduling agent actions over time.
How agents interact with one another and with their environment.
How the components come together to produce a complete agent-based model.
Before we look at each part in detail, try a quick self-check against the five you have just named. In the activity below, each description names one of the five elements. Match every description to the right element, and see how the parts fit together.
Key takeaways#
Key takeaways
Every ABM combines agents, an environment, time, interactions, and a model that coordinates the whole system.
The five elements form a reusable checklist for reading, designing, and debugging any agent-based model.
A model’s system-level behaviour depends on how these elements are connected, not on any one element in isolation.
These same five elements return in every model in this book, so it is worth being able to name them on sight. Begin with 2.2.1: Agents, which examines the agents themselves.