Concurrent Multi-agent Planning for Emergent Narrative

Abstract

This thesis builds upon a prior work on emergent behaviour, which provided a system to easily create a turn-based environment in which agents show emergent behaviour using only a very simple agent definition. As a fist step we integrate this system into Unity and design a small and simple simulation to show that it can easily be integrated into “real-world” systems. The main contribution is then the extension of the prior system such that it is no longer only limited to turn-based applications but can also be used for real-time simulations where multiple agents can plan and act at the same time, which makes it much more flexible and also more viable for real applications.