Open Access

REVIEW

A Review of Next-Generation Smart Manufacturing Enabled by Engineering Systems, Materials Modeling, and High-Performance Computing: A System-Oriented Perspective for Semiconductor Manufacturing

Hsiao-Chun Han¹, Der-Chen Huang^2,*, Chin-Ling Chen^3,*

1 Department of Applied Artificial Intelligence, Ming Chuan University, Taoyuan City, Taiwan
2 Department of Computer Science and Engineering, National Chung Hsing University, Taichung City, Taiwan
3 Department of Computer Science and Information Engineering, Chaoyang University of Technology, Taichung City, Taiwan

* Corresponding Authors: Der-Chen Huang. Email: ; Chin-Ling Chen. Email:

Computers, Materials & Continua 2026, 89(1), 10 https://doi.org/10.32604/cmc.2026.084216

Received 18 April 2026; Accepted 22 June 2026; Issue published 13 August 2026

Abstract

Semiconductors represent the most complex production activities and stand at the forefront of smart manufacturing. In particular, the yield of advanced processes is strongly influenced by coupling among engineering systems, material behavior, and computational infrastructure. Consequently, the integration of deep learning (DL) and digital twins has become essential for driving the next-generation transformation of smart manufacturing. However, existing reviews predominantly organize literature through algorithm-oriented taxonomies, while isolated AI paradigms alone remain insufficient to effectively capture system-level interactions and industry-driven technological evolution. Therefore, this study proposes a system-oriented and industry-driven review framework, termed the System-under-Industry Guided Literature Review (S_IGLR), to reorganize existing literature. Specifically, 103 survey papers and the 13,377 studies they reference are systematically analyzed. In addition, a triadic ESA–MM–HPC framework is introduced as a system-oriented conceptual representation for analyzing next-generation semiconductor smart manufacturing systems, capturing the coupling among engineering system analysis (ESA), material modeling (MM), and high-performance computing (HPC) in complex manufacturing environments. Furthermore, a conceptual modeling perspective is proposed for predictive digital twin (DT) analysis. Overall, this study uses the semiconductor industry as a representative domain to extend existing smart manufacturing research from algorithm-centered approaches toward more system- and structure-oriented perspectives. The perspectives provide a unified, scalable foundation for AI-driven industrial systems research, advancing manufacturing technologies and discussing challenges associated with globally distributed semiconductor manufacturing and industrial reshoring.

---

Keywords

---