A novel approach to cognitive automation of assembly processes is introduced. An experimental assembly cell with two robots has been designed to proof the concept. The cell’s numerical control – termed a cognitive control unit (CCU) – is able to simulate human information processing at a rule-based level of cognitive control on the basis of the SOAR cognitive architecture. Thus the CCU can plan assembly processes autonomously and can react to changes in assembly processes due to increasing number of products that have to be assembled in a large variety in production space as well as changing or uncertain conditions. To develop a “Humanoid-Mode” for automated assembly systems similar to the H-metaphor for automated vehicles human assembly strategies where identified in empirical investigations and formulated as production rules. When the CCU is enriched with these production rules underlying human heuristics, a significant increase of the predictability of a robot when assembling products can be achieved.