Open Access

REVIEW

A Review of the Application of Machine Learning in Additive Manufacturing

Yuyin Ma¹, Yufang Liu¹, Yijun Lu², Zhen Tian³, Fujiang Yuan⁴, Yanhong Peng^4,*

1 Xinjiang Key Laboratory of Intelligent Computing and Smart Applications, School of Software, Xinjiang University, Urumqi, China
2 Department of Computer Science and Engineering, Waseda University, Tokyo, Japan
3 James Watt School of Engineering, University of Glasgow, Glasgow, UK
4 College of Mechanical Engineering, Chongqing University of Technology, Chongqing, China

* Corresponding Author: Yanhong Peng. Email:

(This article belongs to the Special Issue: Computational Materials Design and Intelligent Processing for Advanced Alloys and Manufacturing Systems)

Computers, Materials & Continua 2026, 88(2), 2 https://doi.org/10.32604/cmc.2026.080309

Received 13 February 2026; Accepted 17 April 2026; Issue published 15 June 2026

Abstract

Additive manufacturing (AM) has emerged as a transformative technology in modern manufacturing, offering unprecedented capabilities for producing complex geometries and customized components. However, the widespread adoption of AM is hindered by insufficient quality control, stemming from the multi-factor coupling characteristics of the manufacturing process. Machine learning (ML) presents a promising solution by enabling data-driven approaches to process optimization, quality prediction, and defect detection. This review examines the application landscape of ML techniques in AM through comprehensive analysis of recent literature. The study categorizes ML applications into four primary domains: real-time process monitoring and control, process parameter optimization and prediction, material property prediction, and quality inspection and defect identification. Random Forest (RF), neural networks (NN), and Support Vector Machines (SVM) emerge as the most widely adopted algorithms, demonstrating strong performance in handling high-dimensional, nonlinear relationships between process parameters and product quality. The analysis reveals that while ML methods have achieved significant success in offline prediction tasks, most research remains at the supervised learning stage, lacking cross-material adaptability, real-time feedback control capabilities, and model interpretability. The review identifies critical gaps in current research, including the need for closed-loop autonomous control systems, transfer learning across different materials and machines, and physics-informed ML models that integrate domain knowledge. This work provides a comprehensive reference for researchers and practitioners, highlighting both the achievements and limitations of ML applications in AM, and proposing future directions toward intelligent, autonomous, and high-reliability AM systems.

---

Keywords

---