KARMEN: A Knowledge Graph Based Proposal to Capture Expert Designer Experience and Foster Expertise Transfer

[Context] At the cusp of Industry 4.0 and against a backdrop of fierce competition, manufacturing companies must design and manufacture increasingly complex and cost-effective products. Human resources must therefore preserve and maintain their knowledge and the intellectual heritage of their experts. [Problem] In the next few years, there will be a lack of skilled resources in the manufacturing industry due to retirements. Let’s also mention the turnover of consultants working within these companies. It is essential to implement solutions today in order to protect the intellectual heritage of tomorrow. This paper ambition to answer to how can the knowledge of these experts be captured and used, and how knowledge graph could be a suitable tool to achieve this objective. [Proposal] This article proposes a methodology for implementing KBE (Knowledge Based Engineering) solutions. This methodology called KARMEN (Knowledge Access Request for Manufacturing and Engineering by Network graph) is based on an FBS type ontology (Function, Behavior, Structure) as well as on the exploitation of Knowledge Graphs. A use case of redesigning a mechanical part for metal additive manufacturing will be presented. Besides, an experimental protocol will be specified to capture the knowledge of business experts within a graph-oriented database built on Neo4J. Finally, it will demonstrate that navigation within a knowledge graph can be a powerful tool for knowledge transfer and support in designing novice profile.