Read the full-text of this article for free

Title: Reproducible decision support for industrial decision making using a knowledge extraction platform on multi-objective optimisation data

Authors: Simon Lidberg; Amos H.C. Ng

Addresses: School of Engineering Science, University of Skövde, Högskolevägen, Box 408, 541-28, Skövde, Sweden; Manufacturing Engineering Development, Volvo Group Trucks Operations, John G. Grönvalls Plats 10, 541-37, Skövde, Sweden ' School of Engineering Science, University of Skövde, Högskolevägen, Box 408, 541-28, Skövde, Sweden

Abstract: Simulation-based optimisation enables companies to take decisions based on data, and allows prescriptive analysis of current and future production scenarios, creating a competitive edge. However, effectively visualising and extracting knowledge from the vast amounts of data generated by many-objective optimisation algorithms can be challenging. We present an open-source, web-based application in the R language to extract knowledge from data generated from simulation-based optimisation. For the tool to be useful for real-world industrial decision-making support, several decision makers gave their requirements for such a tool. This information was used to augment the tool to provide the desired features for decision support in the industry. The open-source tool is then used to extract knowledge from two industrial use cases. Furthermore, we discuss future work, including planned additions to the open-source tool and the exploration of automatic model generation. [Submitted 13 November 2022; Accepted 29 March 2023]

Keywords: knowledge-extraction; reproducible science; simulation-based optimisation; industrial use-case; decision-support; knowledge-driven optimisation.

DOI: 10.1504/IJMR.2023.135645

International Journal of Manufacturing Research, 2023 Vol.18 No.4, pp.454 - 480

Received: 13 Nov 2022
Accepted: 29 Mar 2023
Published online: 20 Dec 2023 *