Page 29 - IJAMD-2-3

International Journal of AI for
Materials and Design AI applications in composite materials

such as VAEs and GANs. Automation models, which (NRF), funded by the Ministry of Science and ICT (RS-
make automated decisions during the manufacturing 2023-00260461).
process through real-time data analysis, prediction, and
optimization, play a significant role in improving efficiency Conflict of interest
and quality, and are expected to make key contributions Seong Su Kim is an Editorial Board Member of this
to enable widespread adoption and popularization of journal but was not in any way involved in the editorial
composite materials. Through various research examples, and peer-review process conducted for this paper, directly
the paper confirms that ML provides a new paradigm for or indirectly. Separately, other authors declared that they
research in composite materials. have no known competing financial interests or personal
Despite these advancements, several critical challenges relationships that could have influenced the work reported
remain. Data scarcity continues to limit ML performance, in this paper.
highlighting the need for data-efficient approaches, Author contributions
such as transfer learning, few-shot learning, and the
development of large-scale open datasets. Enhancing Conceptualization: Hyunsoo Hong, Samuel Kim
data accessibility and promoting widespread data-sharing Visualization: Samuel Kim
within the research community will be highly beneficial Writing – original draft: Hyunsoo Hong
for accelerating ML-driven advancements in composite Writing – review & editing: All authors
materials research. Moreover, to maximize the practical
utilization of ML, combining and applying multiple Ethics approval and consent to participate
ML methodologies appropriately becomes increasingly Not applicable.
important. Generalization issues, particularly the difficulty
of applying trained models to new materials or conditions, Consent for publication
can be addressed by embedding physics-based knowledge Not applicable.
into ML frameworks through hybrid modeling and domain
adaptation techniques. In addition, rather than relying on Availability of data
blind data accumulation, more efficient learning through No datasets were generated or analyzed during the current
intentional human-driven data curation is expected to study.
enhance ML performance. Challenges regarding training
instability and lack of interpretability require further References
research on regularization and the adoption of XAI
approaches. Furthermore, the real-world deployment 1. Silver D, Schrittwieser J, Simonyan K, et al. Mastering
the game of go without human knowledge. Nature.
of AI models faces significant challenges related to 2017;550(7676):354-359.
robustness, latency, and safety. These challenges can be
mitigated through advances in edge computing, digital doi: 10.1038/nature24270
twin frameworks, real-time AI inference technologies, 2. Silver D, Huang A, Maddison CJ, et al. Mastering the game
and specialized ML hardware, such as neural processing of Go with deep neural networks and tree search. Nature.
units. Finally, integrating emerging technologies, such as 2016;529(7587):484-489.
quantum computing, may further enhance the scalability, doi: 10.1038/nature16961
reliability, and industrial adoption of AI-driven solutions 3. Brown T, Mann B, Ryder N, et al. Language models are few-
in the research and manufacturing of composite materials. shot learners. Adv Neural Inf Process Syst. 2020;33:1877-1901.
Acknowledgments 4. Bai Y, Kadavath S, Kundu S, et al. Constitutional AI:
Harmlessness From AI Feedback. [arXiv Preprint]; 2022.
None.
5. Grigorescu S, Trasnea B, Cocias T, Macesanu G. A survey
Funding of deep learning techniques for autonomous driving. J Field
Robot. 2020;37(3):362-386.
This research was supported by the Nano & Material
Technology Development Program through the National doi: 10.1002/rob.21918
Research Foundation of Korea (NRF), funded by the 6. Holzinger A, Keiblinger K, Holub P, Zatloukal K, Müller H.
Ministry of Science and ICT (No. RS-2024-00450477). AI for life: Trends in artificial intelligence for biotechnology.
This work was supported by the National R&D Program N Biotechnol. 2023;74:16-24.
through the National Research Foundation of Korea doi: 10.1016/j.nbt.2023.02.001

Volume 2 Issue 3 (2025) 23 doi: 10.36922/IJAMD025210016