A Multi-agent Potential Field based bot for a Full RTS Game Scenario

Document type: Conference Papers
Peer reviewed: Yes
Full text:
Author(s): Johan Hagelbäck, Stefan J. Johansson
Title: A Multi-agent Potential Field based bot for a Full RTS Game Scenario
Conference name: Artificial Intelligence and Interactive Digital Entertainment (AIIDE)
Year: 2009
Publisher: Association for the Advancement of Artificial Intelligence
City: Stanford, USA
Organization: Blekinge Institute of Technology
Department: School of Computing (Sektionen för datavetenskap och kommunikation)
School of Computing S-371 79 Karlskrona
+46 455 38 50 00
Authors e-mail: johan.hagelback@bth.se, stefan.johansson@bth.se
Language: English
Abstract: Computer games in general, and Real Time Strategy games in particular is a challenging task for both AI research and game AI programmers. The player, or AI bot, must use its workers to gather resources. They must be spent wisely on structures such as barracks or factories, mobile units such as soldiers, workers and tanks. The constructed units can be used to explore the game world, hunt down the enemy forces and destroy the opponent buildings. We propose a multi-agent architecture based on artificial potential fields for a full real time strategy scenario. We validate the solution by participating in a yearly open real time strategy game tournament and show that the bot, even though not using any form of path planning for navigation, is able to perform well and win the tournament.
Subject: Computer Science\Artificial Intelligence
Digital Game Development\General
Keywords: Agents, Swarm Intelligence and Emergent Behavior, Multidisciplinary Topics and Applications, Computer Games