Open this publication in new window or tab >>
2026 (English) In: Metals, ISSN 2075-4701, Vol. 16, no 9, article id 1043Article in journal (Refereed) Published
Abstract [en] Despite advances in digital engineering tools, early-stage sheet metal forming tooling design remains strongly dependent on the tacit knowledge of experienced engineers, making systematic knowledge reuse difficult. This study addresses the need for a structured method to capture, organize, retrieve, and reuse historical tooling knowledge during early design decision-making. A data-driven decision-support framework is proposed, in which previous tooling projects are represented as structured knowledge cases containing CAD-derived features and performance-related data. The framework is developed within a Case-Based Reasoning paradigm and consists of three modules: a Difficulty Assessment Module for estimating manufacturing difficulty, a Similarity Retrieval Module for identifying comparable historical cases, and a Performance Module for linking design decisions with maintenance and operational outcomes. A prototype implementation was developed, tested, and evaluated through expert validation workshops with industrial tooling partners. The results indicate that the proposed framework has the potential to support more interpretable and systematic early-stage tooling decisions by formalizing design knowledge and enabling the retrieval of relevant past cases. The study concludes that data-driven knowledge reuse can reduce reliance on individual experience, may potentially reduce the design lead time, and support early design stages.
Place, publisher, year, edition, pages
MDPI, 2026
Keywords sheet metal tooling, engineering design, AI, ML, knowledge reuse, early design phase, data-driven design
National Category
Production Engineering, Human Work Science and Ergonomics
Identifiers urn:nbn:se:hj:diva-73931 (URN) 10.3390/met16091043 (DOI) 001891219100001 () 2-s2.0-105051788524 (Scopus ID) GOA;;73931 (Local ID) GOA;;73931 (Archive number) GOA;;73931 (OAI)
Note This research was supported by the Smart Tooling project, funded by Stiftelsen Seydlitz MP bolagen with grant number 305.
2026-09-282026-09-282026-10-13 Bibliographically approved