The DiDIY-Factory model is a model of an abstract factory. Its purpose is to investigate the impact Digital Do-It-Yourself (DiDIY) could have on the domain of work and organisation.

DiDIY can be defined as the set of all manufacturing activities (and mindsets) that are made possible by digital technologies. The availability and ease of use of digital technologies together with easily accessible shared knowledge may allow anyone to carry out activities that were previously only performed by experts and professionals. In the context of work and organisations, the DiDIY effect shakes organisational roles by such disintermediation of experts. It allows workers to overcome the traditionally strict organisational hierarchies by having direct access to relevant information, e.g. the status of machines via real-time information systems implemented in the factory.

A simulation model of this general scenario needs to represent a more or less abstract manufacturing firm with supervisors, workers, machines and tasks to be performed. Experiments with such a model can then be run to investigate the organisational structure –- changing from a strict hierarchy to a self-organised, seemingly anarchic organisation.

The objective of this agent-based model is to test different language education orientations and their consequences for the EU population in terms of linguistic disenfranchisement, that is, the inability of citizens to understand EU documents and parliamentary discussions should their native language(s) no longer be official. I will focus on the impact of linguistic distance and language learning. Ideally, this model would be a tool to help EU policy makers make informed decisions about language practices and education policies, taking into account their consequences in terms of diversity and linguistic disenfranchisement. The model can be used to force agents to make certain choices in terms of language skills acquisition. The user can then go on to compare different scenarios in which language skills are acquired according to different rationales. The idea is that, by forcing agents to adopt certain language learning strategies, the model user can simulate policies promoting the acquisition of language skills and get an idea of their impact. In this way the model allows not only to sketch various scenarios of the evolution of language skills among EU citizens, but also to estimate the level of disenfranchisement in each of these scenarios.

MASTOC-LLM extends the classic Multi-Agent System Tragedy of the Commons (MASTOC) model by replacing hard-coded behavioral rules with autonomous decision-making powered by large language models (LLMs). Three heterogeneous agents manage herds of cows on a shared grassland commons. Each tick, an agent receives a structured prompt describing current resource levels, its own herd size, peer behavior, and — optionally — a rolling memory of recent rounds and messages from neighboring agents. The LLM returns a stocking decision (add, remove, or hold cows) together with a natural-language rationale and, when communication is enabled, a short message to broadcast to peers.

The model is designed to test whether LLM agents spontaneously develop Ostrom-style common-pool resource governance (mutual monitoring, graduated sanctions, graduated rule revision) or instead fall into identifiable failure modes. Preliminary experiments with Claude Haiku 4.5, GPT-5.4-mini, and DeepSeek R1:32b have revealed four recurring collapse patterns — Cooperative Paralysis, Defection Cascade, Overshoot-Panic, and Hybrid Architecture Failure — whose onset timing is sensitive to memory length, inter-agent communication, and the post-training alignment approach of the underlying model.

MASTOC-LLM is intended as a laboratory for generative agent-based modelling (GABM) methodology: it provides a clean, well-understood commons baseline against which LLM behavioral hypotheses can be systematically tested and compared across models, parameter sweeps, and alignment regimes.

This repository contains the Python implementation of an agent-based model investigating how localized boundary-crossing dynamics generate large-scale connectivity in structured multi-attractor landscapes.

Agents evolve in a continuous two-dimensional environment composed of attractor basins. A fraction of agents exhibits exploratory higher-mobility dynamics, while the remaining agents remain locally constrained. The model analyzes how localized configurational transitions accumulate into transition networks that progressively integrate the explored state space.

The repository includes:

…

Micro-targeted vs stochastic political campaigning agent-based model simulation. Written by Toby D. Pilditch (University of Oxford, University College London), in collaboration with Jens K. Madsen (University of Oxford, London School of Economics)

The purpose of the model is to explore the various impacts on voting intention among a population sample, when both stochastic (traditional) and Micto-targeted campaigns (MTCs) are in play. There are several stages of the model: initialization (setup), campaigning (active running protocols) and vote-casting (end of simulation). The campaigning stage consists of update cycles in which “voters” are targeted and “persuaded” - updating their beliefs in the campaign candidate / policies.

Agriculture is the largest water-consuming sector worldwide, responsible for almost 70% of the world’s total freshwater consumption. Agricultural water reuse is one of the most sustainable and reliable methods to alleviate water shortages worldwide. However, the dynamics of agricultural water reuse adoption by farmers and its impacts on local water resources are still unknown to the scientific community, according to the literature. Therefore, the primary purpose of the WRAF model is to investigate the micro-level dynamics of agricultural water reuse adoption by farmers and its impacts on local water resources. The WRAF was developed using agent-based modeling as an exploratory tool for scenario analysis. The model was specifically designed for researchers and water resources decision-makers, especially those interested in natural resources management and water reuse.
WRAF simulates a virtual agricultural area in which several autonomous farms operate. It also simulates these farms’ water consumption dynamics. The developed model includes two types of agents: farmers and wastewater treatment plants. In general, farmer agents are the main water-consuming agents, and wastewater treatment plant agents are recycled water providers in the WRAF model. Dynamic simulation of agricultural water supply and demand in the area allows the user to observe the results of various irrigation water management scenarios, including recycled water. The models also enable the user to apply multiple climate change scenarios, including normal, moderate drought, severe drought, and wet weather conditions.

The purpose of the model is to simulate the spatial dynamics of potato late blight to analyse whether resistant varieties can be used effectively for sustainable disease control. The model represents an agricultural landscape with potato fields and data of a Dutch agricultural region is used as input for the model. We simulated potato production, disease spread and pathogen evolution during the growing season (April to September) for 36 years. Since late blight development and crop growth is weather dependent, measured weather data is used as model input. A susceptible and late blight resistant potato variety are distinguished. The resistant variety has a potentially lower yield but cannot get infected with the disease. However, during the growing season virulent spores can emerge as a result of mutations during spore production. This new virulent strain is able to infect the resistant fields, resulting in resistance breakdown. The model shows how disease severity, resistance durability and potato yield are affected by the fraction of fields across a landscape with a disease-resistant potato variety.

We develop an IBM that predicts how interactions between elephants, poachers, and law enforcement affect poaching levels within a virtual protected area. The model is theoretical at this stage and is not meant to provide a realistic depiction of poaching, but instead to demonstrate how IBMs can expand upon the existing modelling work done in this field, and to provide a framework for future research. The model could be further developed into a useful management support tool to predict the outcomes of various poaching mitigation strategies at real-world locations. The model was implemented in NetLogo version 6.1.0.

We first compared a scenario in which poachers have prescribed, non-adaptive decision-making and move randomly across the landscape, to one in which poachers adaptively respond to their memories of elephant locations and where other poachers have been caught by law enforcement. We then compare a situation in which ranger effort is distributed unevenly across the protected area to one in which rangers patrol by adaptively following elephant matriarchal herds.

WWHW is an agent-based model designed to allow the exploration of the emergence, resilience and evolution of cooperative behaviours in hunter-fisher-gatherer societies.

Patagonia PSMED is an agent-based model designed to study a simple case of Evolution of Ethnic Differentiation. It replicates how can hunter-gatherer societies evolve and built cultural identities as a consequence of the way they interacted.