High-level behavior programming

Reapplication of the same agent behavior architecture in a different domain usually requires reimplementation of the architecture framework such as its behavior management code. This requires extensive knowledge and manipulation of the native code, which is time consuming and error prone. We conjecture that it will be beneficial to separate the process of designing agent behavior architecture for high-level behaviors and incorporating the architecture to a given agent platform with implemented platformdependent atomic behaviors. Towards that goal, we provide an intuitive and flexible high-level behavior-based language (HLBL) that can be used to describe easily the architecture for an agent to facilitate the designing process, and a code generation system (B2C) to automate the integrating process.