Abstract:
This paper presents a theory of autonomous action that combines the perspectives of three disciplines: neuropsychology, AI planning, and control system technology. Each discipline brings complementary contributions to the common goal of understanding the nature of autonomous action, and the combined theory requires extensions to each of the component theories. Our overall goal is to: Extend AI planning methods to perform the functions of the neuropsychological models within the context of real-time, closed-loop control applications.