Reasoning about BDI Agents from a Programming Languages Perspective

In this paper, we summarize an approach to reasoning about a class of BDI agent architectures based on PRS. The theory is formalized using a logic, Agent Dynamic Logic (ADL), that combines elements from Computation Tree Logic, Propositional Dynamic Logic and Rao and Georgeff’s BDI Logic. The motivation of this work is to develop a logical framework that is at once rigorous in providing formal notions of belief, desire and intention, yet which is also computationally grounded in the operational behaviour of this architecture, so as to enable formal reasoning about the behaviour of agents in this class. We illustrate the model theory with a simple “waypoint following” agent.