An Embodied Mechanism for Autonomous Action Selection and Dynamic Motivation

Neural schema mechanism is an embodied autonomous agent control structure that makes use of both neural network and symbolic constructs to learn sensory motor correlations and abstract concepts through its own experience. The mechanism can also learn which intermediate states or goals should be achieved or avoided based on its primitive drives. In addition, a psychological theory of consciousness is modeled that allows the system to come up with creative action sequences to achieve goals even under situations of incomplete knowledge. The result is an architecture for robust action selection that learns not only how to achieve primitive drives, but also learns appropriate sub-goals that are in service of those drives.