A theory of intelligent machine systems

Abstract

A theoretical model is proposed consisting of six basic elements: actuators, sensors, sensory processing, world modeling, behavior generation and value judgment. These elements are integrated into a hierarchical system architecture wherein control bandwidth decreases about an order of magnitude from level to level, goals expand in scope and planning horizons expand in space and time about an order-of-magnitude from level to level, and perception of spatial and temporal patterns, models of the world, and memories of events decrease in resolution and expand in spatial and temporal range by about an order-of-magnitude from level to level. At each level, functional modules perform behavior generation, world modeling, sensory processing, and value judgment. Behavior generation modules perform task planning and execution. The world model is the intelligent system's best estimate of the external world. Value judgments provide an evaluation of hypothesized plans, and perceived objects, events, and situations. Sensory processing modules transform sensory maps into world model maps and extract entity attributes and states. Sensory feedback control loops are closed at every level.<>