Tutorial "Creating game bots in a few easy steps"

• what controls virtual characters (bots) in computer games?
• how they can find a path in a complex virtual environment automatically?
• what finite-state machines, reactive rules and other fancy terms really mean?
• whether bots can have emotions and how to make them laugh and cry?
• how they can replay a given scenario, yet interact with a human user at the same time?

Do you want to program them on your own on your home PC?

Creating gaming AI for the first time is always difficult. Our toolkit Pogamut facilitates first steps with control mechanisms of virtual characters without burdening the beginner with complex programming issues that surround this topic (data structures, engine programming etc.).

Pogamut is an open source toolkit for rapid development of behaviour for virtual characters embodied in a 3D environment of the Unreal Tournament 2004 computer game intentionally created to facilitate gaming AI programming to newcomers. Although connected to UT, it can be used to start with general characters’ AI, e.g., for RPGs or even artistic applications and cognitive modeling. Pogamut is also able to support users advancing by providing additional data structures and programming primitives. On the one hand, it allows for rapid creation of simple rule-based bots, on the other hand, it facilitates development of challenging bots exploiting fancy AI techniques, such as neural networks, evolutionary programming, fuzzy logic or planning.

In this tutorial, the basics of gaming AI will be introduced and a simple bot will be created using Pogamut. Participants can install Pogamut on their notebooks and develop their first bots during the tutorial.

After visiting this tutorial, the participants will know:

• how basic gaming AI mechanisms work in theory
• how they can be coded using Pogamut, i.e. how these mechanisms work in practise
• how Pogamut can be used in research and education

Anyone who has at least moderate IT knowledge and wants to know how to create bots. We will start at the beginning – no previous gaming AI knowledge is assumed.

In particular, computer science, cognitive science, new media and art students and researchers are welcomed.

Additionally, the tutorial addresses teachers of gaming AI looking for a toolkit in which students can practice development of virtual characters.

Figure: An example of two interacting bots (Copyright © 2009 Epic Games 2009)