Abstract:
This paper describes an architecture for controlling autonomous agents, building on previous work addressing reactive and deliberative control methods. The proposed multi-layered architecture allows a resource-bounded, goal-directed agent to reason predictively about potential conflicts by constructing causal theories which explain other agents’ observed behaviors and hypothesize their goals and intentions; at the same time it enables the agent to operate autonomously and to react promptly to changes in its real-time physical environment. A number of criteria which influenced the design and implementation of the architecture, in particular its action control component, are also discussed.