Efficient reuse and variability management for cyber-physical production system families

Abstract

Today, the demand for highly individualized products, such as cars, increases product variants and drives flexible and reconfigurable manufacturing. Cyber-Physical Production Systems (CPPSs), like automated car factories, are software-intensive systems using modern techniques. They contain production resources, like robots, executing processes such as welding car parts manufacturing customizable products. During CPPS engineering, domain experts aim to systematically and efficiently reuse artifacts, building CPPS families with common and variable features, like work lines with similar resources. However, CPPS engineering involves multiple disciplines, such as mechanical and electrical engineering, creating challenges like discipline-specific implicit knowledge and inefficient, manual processes. However, separating discipline-specific concerns is crucial for agile CPPS engineering. This thesis addresses these challenges with a multifaceted approach. It aims to establish explicit production knowledge models to support multidisciplinary CPPS knowledge exchange. It explores designing methods, reference models, and design patterns for advanced engineering applications to enhance communication and coordination. It further develops integrated reuse and variability management to improve CPPS family design efficiency.The thesis contributes several models and methods. The superimposed Product-Process-Resource (PPR) model visually represents PPR with variability. The PPR meta-model formalizes this repre- sentation, while the PPR Domain-Specific Language externalizes implicit knowledge. For advanced applications, the Industrie 4.0 Asset Network meta-model and I4AN reference model improve coordination. Basic design patterns address recurring engineering problems. The Capability & Skill Reuse framework facilitates systematic reuse. The Extended Iterative Process Sequence Exploration (EIPSE) approach manages the integrated variability that PPRs spawn in CPPS families. This way, it addresses the essential three dimensions of CPPS variability including production process sequences. Evaluation through mixed methods, including domain analyses and case studies, showed effective externalization of knowledge, improved communication, and efficient configuration of CPPS variability. The EIPSE enables the reuse of knowledge and artifact management, reducing design effort for CPPS families.