What is it about?

The paper proposes and architecture that enables the simulation of human-like strategic intelligence in real-time strategy games. It uses multiple research fields to accomplish this, namely multi-agent architectures, holonic multi-agent architectures, the Beliefs-Desires-Intentions (BDI) model and holonic superposition intelligent agents to achieve this. The paper also provides an experiment with results to determine how effective this architecture is.

Featured Image

Why is it important?

Simulating human-like strategic intelligence in real-time strategy (RTS) games is quite difficult. Either the AI is too predictable, or too difficult. However, this research paper attempts to show how multiple paradigms can work seamlessly to create more realistic human-like strategic intelligence in RTS games. This is also important for systems that may need to implement very complex, strategic thinking in non game systems.


I hope this article finds it way to people that wish to delve deeper into the AI prospects of RTS games. RTS games are quite challenging as all constraints, conditions and rules have to be constantly evaluated and monitored and run-time. I also hope that this paper demonstrates how multi-agent systems, and especially holonic multi-agent systems is an adequate approach to monitor, evaluate and respond to real-time triggers/constraints.

Gerard Gouws
University of Johannesburg

Read the Original

This page is a summary of: Simulating Human Strategic Vision in Real-Time Strategy Games with Holonic Superposition Intelligent Multi-Agent Systems, Journal of Integrated Design and Process Science, January 2020, IOS Press, DOI: 10.3233/jid190005.
You can read the full text:



The following have contributed to this page