Multi-Agents in a Virtual Regional Landscape

Abstract

The Virtual Region More was a research project to investigate how to utilize visualization and simulation methods as planning tools for research and teaching in the More region on the west coast of Norway. The project was developed using a bottom-up process to integrate multi-agents in large 3D terrain models. The Virtual Region More became an arena for testing agent models in shipping, fish farming, ecosystems, virus swarms and energy production. The most important result of the project was recognizing that simulating multi-agent systems in position-dependent landscapes leads to complex landscape-dependent systems. Multi-agents in large landscapes need generic methods to modulate generic agents. Generic multi-agents in large 3D landscapes need a system view to represent multi-agents in organizations, where agents are controlled by cost functions.