Page 8 - IJAMD-2-2

International Journal of AI for
Materials and Design AI-driven material development for AM

machining, thereby significantly improving material computational design and experimental realization. The
utilization efficiency. As a general term encompassing increasing research interest in this field is reflected in
2
various techniques, AM has been adopted across the growing number of AI-driven AM publications over the
multiple industries, including aerospace, biomedical, past decade, as illustrated in Figure 1.
and automotive sectors. Each AM technique, whether Given the growing importance of AI in materials
designed for metals, polymers, or bioinks/biomaterial discovery for AM, this review aims to provide a critical
inks, operates under distinct processing principles that assessment of recent advancements in AI-driven material
influence the manufacturability and performance of development across different AM material categories.
fabricated components. 3,4 Unlike many existing reviews that predominantly focus
Designing and optimizing materials for AM remains a on in situ monitoring, process control, or the printability
challenging task, as it has traditionally relied on heuristic- assessment of feedstock materials using AI, this review
9
driven trial-and-error methods that are both time- specifically emphasizes AI-driven material development.
consuming and resource-intensive. These conventional We adopt a narrow definition of material development,
approaches struggle to efficiently navigate the complex which centers on the design and optimization of material
and non-linear relationships between processing properties or compositions, excluding applications
conditions, microstructure evolution, and resulting solely related to manufacturability, defect detection, or
material properties. Moreover, most existing materials geometry control. This focused perspective highlights
were originally developed for conventional manufacturing the unique role of AI in accelerating the discovery and
techniques and are often not suitable for AM, leading to tailoring of materials for AM. The framework of the
issues such as poor printability, defect formation, and review is presented in Figure 2. The discussion covers
variability in mechanical performance. The growing need data collection and pre-processing methods, AI-enabled
5,6
for AM-specific materials with tailored functionalities material development strategies, and key applications in
demands a paradigm shift in how materials are discovered AM. By summarizing recent progress and highlighting
and developed. In this context, artificial intelligence (AI) existing challenges, this review seeks to offer insights
offers powerful capabilities to handle high-dimensional into the evolving intersection of AI and AM materials
datasets, identify hidden patterns, and accelerate the development, paving the way for future innovations in
design process by predicting material performance based this field.
on compositional and process parameters. By moving
beyond empirical heuristics, AI-driven methods enable 2. Overview of various AM techniques and
more systematic and scalable exploration of the vast design the need for material development
space, offering a promising alternative to traditional There are various AM techniques that enable the
approaches. fabrication of complex structures. Each AM technique

In recent years, AI has demonstrated remarkable success imposes distinct material requirements, influencing
in media applications based on image and voice big data, processing feasibility and final part performance. This
and it has recently expanded into diverse fields, such as section provides a concise description of seven key AM
medicine, education, and engineering. Its ability to techniques (Figure 3), their material requirements, and
7,8
analyze vast datasets, identify complex patterns, and make challenges in material development.
predictive decisions has also positioned AI as a powerful
tool in scientific and industrial applications, including 2.1. Vat photopolymerization (VPP)
materials science. In particular, AI has emerged as a The VPP process is an AM technology based on a liquid
transformative tool in accelerating materials development photosensitive resin, which is cured layer by layer using a
for AM. By leveraging data-driven approaches, AI enables light source, resulting in a solid 3D part. Several variations
the prediction of material properties, optimization of alloy of VPP exist, including stereolithography, digital light
compositions, and exploration of new material spaces with processing, two-photon polymerization, and volumetric
reduced experimental effort. AI-driven methods facilitate 3D printing. Commonly employed light sources include
11
the rapid identification of processable materials with tailored laser, digital light projection, and LED systems, offering
properties, improving the efficiency of AM applications varying resolutions and processing speeds. VPP is capable
12
across metals, polymers, and bioinks/biomaterial inks. As of achieving micron- to nanometer-scale resolution,
AM continues to evolve toward more sophisticated and making it well-suited for manufacturing components
customized applications, AI-driven material development requiring high dimensional accuracy and superior
plays an increasingly vital role in bridging the gap between surface quality. This technique has been widely applied
13

Volume 2 Issue 2 (2025) 2 doi: 10.36922/IJAMD025100007