A Knowledge Graph-based Approach for Capturing Implicit Knowledge in Composite Manufacturing

This is the latest version of this item.

Abstract

Expert knowledge plays a vital role in the design and manufacturability of complex engineering components, especially in domains where geometry and process capabilities are tightly coupled, such as in composite manufacturing. This technology is widely used in aerospace applications, particularly in the production of lightweight, high-strength aircraft structures as fuselage sections or wing components. While technical standards provide formal guidelines for manufacturing processes and design rules, much of the process-specific, experience-based knowledge remains implicit and undocumented. To make this implicit knowledge accessible and reusable, we present a structured approach to modeling manufacturing knowledge using knowledge graphs. The proposed methodology combines ontology engineering with structured expert knowledge acquisition. It begins with identifying relevant sources of explicit knowledge, such as industry standards, technical handbooks, and process documentation, and is complemented by the extraction of implicit knowledge through structured interviews and workshops with domain experts. This knowledge is mapped onto a domain-specific ontology that formalizes key entities such as manufacturing processes, materials, manufacturing features and manufacturing restrictions. The resulting knowledge graph enables semantic linking and contextualization of process-specific know-how, making it available for applications such as design assistance, manufacturability analysis or production planning. To demonstrate the applicability of this method, we focus on the domain of composite manufacturing, where production feasibility often depends on geometric details, material behavior, and process-specific limitations. In the context of this use case, we conducted a dedicated expert workshop to identify and document the types of geometries typically supported by the process, as well as recurring challenges encountered during production. Using real-world part scenarios and collaborative group analysis, we extracted qualitative insights that are not typically found in documentation or guidelines. These findings were then structured and integrated into the knowledge graph, enabling a reusable representation of practical manufacturing expertise. Our approach demonstrates how formal and informal knowledge sources can be combined to build domain-specific manufacturing knowledge graphs. And thus, creating a foundation for design decisions and knowledge-based engineering tools.

Available Versions of this Item

• A Knowledge Graph-based Approach for Capturing Implicit Knowledge in Composite Manufacturing. (deposited 08 Oct 2025 08:33) [Currently Displayed]