Modeling Expert Knowledge for Optimal CPPS Resource Selection for a Product Portfolio

Abstract

Optimal selection of resources for a Cyber-Physical Production System (CPPS) has a strong potential to reduce the cost of producing product families. Unfortunately, engineers rarely have specific methods and tools to represent their expert knowledge for applying mathematical optimization approaches. Representing this knowledge requires capabilities to model the dependencies between products, production processes, resources, skills, and their variability. In this paper, we provide these capabilities and focus on improving CPPS resource selection by defining aggregated requirements of a set of similar processes for resources. We illustrate the industrial use case Car Body with Screwed-on Parts on defining requirements for resource candidates suitable for a set of screwing process variants. We introduce the Product-Process-Resource-Skill & Variability (PPRS+V) method for consistently designing aggregated ProductProcess-Resource (PPR) skills required for optimal CPPS resource selection as foundation for engineering support. We introduce the Skill Aggregation algorithm to efficiently aggregate the skills for a process step required to produce a product family. In an initial feasibility study, domain experts found the PPRS+V method and the Skill Aggregation algorithm usable and useful.