Abstraction levels for developing a model of an intelligent decision unit

Technical processes shall be controlled more precisely and therefore require better interconnection. In order to enable machines to make qualified decisions, the interpretation of processes shall happen on a higher level of abstraction. Artificial intelligence (AI) and Cognitive Science (CS) have delivered excellent results so far, but they are still far away from being able to design intelligent units with abilities similar to those of the human mental apparatus. We show in this paper that a new approach in bionics is required using natural science to achieve a breakthrough based on information theory. We discuss the crucial aspects that have directed engineers into a dead end. And we show that new goals can be achieved, if we consider psychoanalysis and establish tight cooperation with experts from this field. The goal is the definition of a formal approach of the lowest abstract level for the mental apparatus.