• Publisher-DOI

During the design of learning factories, the configuration of the technical system plays an important role. The selection of factory elements for learning factories usually takes place based on intuition. Intuitive selection has the disadvantage that the best possible selection is only achieved by chance. In this paper, a method is presented that is based on solving an optimisation problem, which ensures the best possible selection of factory elements considering a target function with restrictions like the budget or the usable area. For this purpose, four steps are distinguished: First, requirements for the technical configuration are derived from the primary goals of the learning factory (step I). Then, factory areas and configuration alternatives containing a combination of factory elements are derived from the products and processes (step II). The utility values of the potential configuration alternatives are determined based on an evaluation method (step III). Subsequently, the best possible combination of configuration alternatives is determined algorithmically by solving an optimisation problem (step IV). This procedure extends the design approach of Abele et al. (2019). It is applied to a learning factory for a company in the mobility industry to evaluate its superiority to an intuitive approach.