A Unifying Metamodel for Skill-Based Manufacturing

This is the replication package for the manuscript “A Unifying Metamodel for Skill-Based Manufacturing” by Alexander Uhl, Jérôme Pfeiffer, Carsten Ellwein, Patrick Fischer, Andrey Morozov, Max Hossfeld, and Andreas Wortmann, which currently is under submission. Files This replication packages contains the following files: full_corpus.csv Corpus of initially collected publications (including snowballing) included_corpus.csv Corpus of included publications data_collection_method.pdf Figure showing the data collection method of the SMS clustering_model_elements.csv Table of clustered model elements of the metamodels found in the SMS Systematic Mapping Study (SMS) Motivation To advance industrial production, researchers are increasingly adopting software-defined manufacturing, a paradigm that decouples manufacturing capabilities from physical hardware—much like mobile operating systems separated app logic from smartphone hardware. Central to this shift is skill-based manufacturing, where production functions are represented as “skills” that define what a system can do rather than how it is physically executed. However, the lack of a unified definition of “skills” across various stakeholders currently hinders the field’s progress. To resolve this ambiguity, this paper presents a systematic mapping study of 531 publications, analyzing how skills are defined, what elements constitute them, and how they are applied in practice. By clarifying the distinctions between skills and capabilities, this study provides a structured foundation to help researchers and practitioners build more flexible, resilient, and evolvable production systems. Research Method With this survey, we answer the following main research questions RQ 1. Who is using skills for which purpose? RQ 2. What is a skill? RQ 3. How are skills realized? RQ 4. How are skills used? with their 16 sub-questions. To this end, we have conducted a systematic mapping study using the literature databases and filtering as illustrated in data_collection_method.pdf. We applied the following criteria for filtering articles: Inclusion criteria Publications in peer-reviewed sources, such as journals, conferences, and workshops. Publications accessible electronically. Publications available in English. Publications where we could deduce from the title, abstract, or keywords that the paper is about skill-based or capability-based manufacturing. Exclusion criteria Publications from sources without systematic peer review processes, such as books, magazines, and websites. Short papers of less than four pages excluding references, such as editorials and reviews. Edited volumes and monographs. Studies presenting literature reviews or mapping studies (discussed as related work). Publications where the term “skill” is referring to human skill in the sense of manual labor or “workers.” Would you like me to convert these lists into a comparison table for easier viewing on your website? Corpus Applying this method has led to a final corpus of 87 publications, which were analyzed to answer the research questions listed above. Clustering of Metamodels Research Method Identify the skill metamodel in the publication. If the skill model is part of a larger system description, determine the boundaries of the skill model within the system. Identify each class in the boundaries of the skill meta-model. Depending on the metamodel, this can be a class (Unified Modeling Language (UML)), a concept (ontology), or a name for a concept (informal description). Try matching the class syntactically and semantically with an existing cluster of previously extracted classes. If no matching cluster exists, create a new cluster and add the extracted class to it. If possible, identify the input and output associations and the cardinality of the clustered class.