Read the full-text of this article for free

Title: Challenges in designing a human-centred AI system in manufacturing

Authors: Yuji Yamamoto; Aranda Muñoz Álvaro; Kristian Sandström

Addresses: School of Innovation, Design, and Engineering, Mälardalen University, Bruksgatan 3B 632 17 Eskilstuna, Sweden ' Research Institutes of Sweden, Stora Gatan 36, Västerås, 72212, Sweden ' School of Innovation, Design, and Engineering, Mälardalen University, Bruksgatan 3B 632 17 Eskilstuna, Sweden

Abstract: Despite successful AI system deployments in manufacturing, methodological support for developing and integrating AI systems into manufacturing processes remains underdeveloped. This paper aims to identify gaps in the methodological support for the early design phase of AI system development in manufacturing. The study reveals the thinking-level challenges that design participants face in the early design phase and identifies remedies for those challenges, which are only superficially addressed in current manufacturing literature. The paper contributes to uncovering the current knowledge gap in developing an actionable methodology for AI system development in manufacturing contexts.

Keywords: human-centred AI; manufacturing; AI system design; machine learning; socio-technical systems; design guidance.

DOI: 10.1504/IJMMS.2024.144289

International Journal of Mechatronics and Manufacturing Systems, 2024 Vol.17 No.4, pp.351 - 369

Received: 21 Feb 2024
Accepted: 10 Sep 2024
Published online: 04 Feb 2025 *