I.I. Naumov1, R.R. Ibadov2, A. E. Merzlikina3

1-3 Don State Technical University (Rostov-on-Don, Russia)
1 naumov-85@yandex.ru; 2 nikulichev@list.ru; 3 leshenko_ant@mail.ru

Selective laser melting (SLM) enables the production of complex, high-quality metal parts, but its widespread adoption is limited by the lack of real-time automated quality control during the printing process. Defects (microcracks, pores, lack of fusion, burnout, and deformation) are typically detected only after the part is completed, resulting in a 35–50% reject rate and significant resource loss. Many critical defects form within the volume or between layers and are not detectable by visual methods, and existing monitoring systems are operator-dependent and do not provide stable, real-time process optimization. As a result, the lack of high-precision layer-by-layer monitoring remains a key barrier to the reliable use of SLM in critical industries.

Objective – to develop an intelligent system for automated optical monitoring of selective laser melting based on machine vision and deep learning, providing layer-by-layer detection and classification of defects in real time with high accuracy and automatic correction of technological printing parameters.

A prototype optical system with hyperspectral, polarization, and RGB cameras processes a SLP layer in 380 ms with a resolution of 4.2 μm. Analysis is performed using three deep learning models (U-Net, Mask R-CNN, and YOLOv8). The ensemble of models achieves an accuracy of over 97% and an F1-score of 96.5%. A dataset of 100 images with 1,000 defects on different materials was used for training. Validation showed 94% defect detection accuracy and a false positive rate of 2–3%. Documents for software and invention registration have been prepared.

The use of PLC technology represents a promising alternative to existing phytoluminaire control systems. The developed solution ensures the reliability required for critical agricultural facilities and meets information infrastructure protection requirements, contributing to the development of sustainable and secure agricultural technologies.

Naumov I.I., Ibadov R.R., Merzlikina A.E. Optical monitoring of selective laser melting. Science Intensive Technologies. 2026. V. 27.
№ 1. P. 29−37. DOI: https://doi.org/ 10.18127/j19998465-202601-03 (in Russian)

1. Yang W., Gan X., He J. Defect identification of 316l stainless steel in selective laser melting process based on deep learning. Processes. 2024. V. 12. № 6. P. 1054.
2. Chen H.Y., Lin C.C., Horng M.H., Chang L.K., Hsu J.H., Chang T.W., Tsai M.C. Deep Learning Applied to Defect Detection in Powder Spreading Process of Magnetic Material Additive Manufacturing. Materials. 2022. V. 15. №16. P. 5662.
3. Mehta M., Shao C. Federated learning-based semantic segmentation for pixel-wise defect detection in additive manufacturing. Journal of Manufacturing Systems. 2022. V. 64. P. 197–210.
4. Caggiano A., Zhang J., Alfieri V., Caiazzo F., Gao R., Teti R. Machine learning-based image processing for on-line defect recognition in additive manufacturing. CIRP annals. 2019. V. 68. № 1. P. 451–454.
5. Luo S., Ma X., Xu J., Li M., Cao L. Deep learning based monitoring of spatter behavior by the acoustic signal in selective laser melting. Sensors. 2021. V. 21. № 21. P. 7179.
6. Fathizadan S., Ju F., Lu Y. Deep representation learning for process variation management in laser powder bed fusion. Additive Manufacturing. 2021. V. 42. P. 101961.
7. Westphal E., Seitz H. A machine learning method for defect detection and visualization in selective laser sintering based on convolutional neural networks. Additive Manufacturing. 2021. V. 41. P. 101965.
8. Manivannan S. Automatic quality inspection in additive manufacturing using semi-supervised deep learning. Journal of Intelligent Manufacturing. 2023. V. 34. № 7. P. 3091–3108.
9. Klippstein S.H., Heiny F., Pashikanti N., Gessler M., Schmid H.J. Powder spread process monitoring in polymer laser sintering and its influences on part properties. Jom. 2022. V. 74. № 3. P. 1149–1157.
10. Cheng L., Jiang Z., Wang H., Ma C., Zhang A., Du H., Zhang Y. Low-rank adaptive transfer learning based for multi-label defect detection in laser powder bed fusion. Optics and Lasers in Engineering. 2025. V. 184. P. 108683.
11. Ouidadi H., Guo S., Zamiela C., Bian L. Real-time defect detection using online learning for laser metal deposition. Journal of Manufacturing Processes. 2023. V. 99. P. 898–910.
12. Lin X., Shen A., Ni D., Fuh J.Y.H., Zhu K. In Situ Defect Detection in Selective Laser Melting Using a Multi-Feature Fusion Method. IFAC-PapersOnLine. 2023. V. 56. № 2. P. 4725–4732.
13. Wang R., Cheung C.F., Wang C. Unsupervised defect segmentation in selective laser melting. IEEE Transactions on Instrumentation and Measurement. 2023. V. 72. P. 1–10.
14. Luo S., Ma X., Xu J., Li M., Cao L. Deep learning based monitoring of spatter behavior by the acoustic signal in selective laser melting. Sensors. 2021. V. 21. № 21. P. 7179.
15. Kolosova Z.A., Kolosova O.P., Shnayder D.A. Obtaining a Set of Vibrational Signals from Rolling Bearings with Varying Degrees of Local Defect Development in the Outer Ring. Advanced Engineering Research (Rostov-on-Don). 2025. V. 25. № 3. P. 242–255.