A Generic Architecture for an Ecosystem of Autonomous Artificial Animals using Dynamic Considerations

Christopher Osmond & Reza Ryan Part of CW17

Using Artificial Intelligence and Non-Player Characters in games has begun to increase rapidly. This is due to both player expectations and availability of new hardware and software technology (Dragert et al, 2012). Artificial Intelligence can increase a player’s immersion and experience with a game as the player see realistic and dynamically reacts to occurrences. However, there is a lack of generic design and implementation of Artificial Intelligence systems that employ more complex algorithms that can be easily integrated and scaled. A common example of a scenario that needs a generic Artificial Intelligence system is an ecosystem of autonomous artificial animals. This research aims to design and implement this system for a simulated virtual forest environment that resembles forest wildlife. This system will employ the Utility AI theory and dynamic considerations to create an ecosystem of autonomous artificial animals. The generic structure makes us enable to scale up our system easily by adding more species in the forest with minimum changes. This design for this system will be shown as well as a walkthrough of the implementation of the system in Unity3D.