Extending AFSIM with Behavioral Emergence

Jeffrey L. Choate

Abstract

The Advanced Framework for Simulation, Integration, and Modeling (AFSIM) provides a capability to evaluate mission level scenarios described in its scripting language. The AFSIM scripting language includes multiple intelligent agent modeling techniques, none of which explicitly provide the ability to have behaviors emerge. Behavioral emergence occurs when a system composed of many simple behaviors working together exhibits a complex pattern not directly attributable to the simpler components. Without behavioral emergence an intelligent agent designer must explicitly write methods for every combination of circumstances that their agent may encounter. A priori consideration of every possible configuration of the world state is intractable. This problem can be solved by adding the Unified Behavior Framework (UBF) to AFSIM which provides a means to explicitly control behavioral emergence. This thesis adds a unified behavior language built on UBF to AFSIM's scripting language and demonstrates behavioral emergence via a case study of these new behaviors in AFSIM.