The need to model and visualise resources for ancient landscapes have prompted an in-depth study into nature-inspired agent-based models of vegetation using concepts from the Science of Complexity and Artificial Life that takes into account diverse interactions and their adaptability to biotic and abiotic factors. The revolutionary modelling technique mimics nature’s emergent phenomenon - global patterns emerge from the local interactions of simple rules embedded within agents. The application of the model on a small terrain in the North Sea Palaeolandscapes Project have subsequently been visualised using games engines.
In view of the greater need to map Doggerland, including the survival and settlement patterns of early hunter-gatherer communities, a pilot study for testing agents as virtual humans is being conducted. At present, initial ground work has been laid for survival and settlement behaviours such as the ability of the agents to discover resources in the landscape and to identify settlement areas based on resources and the suitability of the environment (e.g., proximity to water). The agents have reasoning capability and memories that fade with the passing of time. They are able to identify object ownerships and family or strangers via tagging, build houses, gather food, build fires, burn clearings, and react to environmental changes (e.g., temperature). In the near future agent roles, cooperation, preferences, and culture will be investigated as well as hunting behaviours. The figure below illustrates the agent model with visualisations from the pilot simulation.