MODEL OF SECURE ADAPTIVE CONTROL OF FDM/FFF PRINTING BASED ON MACHINE LEARNING UNDER CYBER THREATS
DOI:
https://doi.org/10.28925/2663-4023.2026.35.1371Keywords:
FDM/FFF printing; machine learning; adaptive control; cybersecurity; cyber-physical system; risk assessment; ResNet; Raspberry Pi 5; control command security, information technologies.Abstract
The article addresses the problem of ensuring secure adaptive control of the FDM/FFF printing process under technological deviations and potential cyber threats to the information and control infrastructure. The growing adoption of additive manufacturing technologies and the integration of 3D printers with network services, embedded computing platforms, and intelligent monitoring tools create a need to improve not only the quality of manufactured products but also the reliability and security of automated control. Particular attention is paid to the timely detection of technological anomalies, prevention of unjustified control interventions, verification of input data reliability, and restriction of unauthorized modifications to printing parameters. The aim of the study is to substantiate a model of secure adaptive control of FDM/FFF printing based on machine learning, which combines technological defect risk assessment, cascaded verification of intelligent analysis results, and generation of control actions considering information security and functional safety requirements. The architecture of an intelligent supervisory control loop that integrates visual data, technological telemetry, and printing task parameters is considered. ResNet-based neural network models are employed for defect identification, while cascaded verification, temporal smoothing of risk estimates, and a hysteresis mechanism are applied to improve the reliability of control decisions. In the context of cybersecurity, mechanisms for access control, G-code command validation, configuration integrity monitoring, and event logging are considered. Their role in limiting undesirable control interventions and maintaining controlled operation of a Raspberry Pi 5-based hardware and software system is identified. Existing experimental results obtained by comparing different configurations of intelligent FDM/FFF printing control systems are analyzed. The evaluated configurations are characterized by defect detection accuracy, false activation rates, control action latency, and the proportion of defective printed products. The comparison demonstrates the potential to reduce technological losses through intelligent monitoring and adaptive intervention mechanisms. At the same time, evaluating the system's resilience against targeted cyberattacks is identified as a separate direction for future research. The practical significance of the considered model lies in its potential application to the development of intelligent control systems for 3D printers with open software interfaces, particularly those based on Marlin and Klipper firmware, under limited computational resources and the need to ensure secure interaction between software components.
Downloads
References
He, H., Zhu, Z., Zhang, Y., Zhang, Z., Famakinwa, T., Yang, R. (2024). Machine condition monitoring for defect detection in fused deposition modelling process: A review. The International Journal of Advanced Manufacturing Technology, 132, 3149–3178. https://doi.org/10.1007/s00170-024-13630-8
Kozhay, K., Turarbek, S., Asselbekova, T., Ali, H., Shehab, E. (2024). Convolutional neural network-based defect detection technique in FDM technology. Procedia Computer Science, 231, 119–128. https://doi.org/10.1016/j.procs.2023.12.183
Werkle, K. T., Trage, C., Wolf, J., Möhring, H.-C. (2024). Generalizable process monitoring for FFF 3D printing with machine vision. Production Engineering, 18, 593–601. https://doi.org/10.1007/s11740-023-01234-2
De La Rosa, A., Armani, A., Golmohamadi, M. (2024). Defect detection and closed-loop feedback using machine learning for fused filament fabrication. Procedia CIRP, 126, 603–608. https://doi.org/10.1016/j.procir.2024.08.247
Brion, D. A. J., & Pattinson, S. W. (2022). Generalisable 3D printing error detection and correction via multi-head neural networks. Nature Communications, 13, Article 4654. https://doi.org/10.1038/s41467-022-31985-y
Kumar, M., Epiphaniou, G., Maple, C. (2025). Security of cyber-physical additive manufacturing supply chain: Survey, attack taxonomy and solutions. Computers & Security, 157, Article 104557. https://doi.org/10.1016/j.cose.2025.104557
Alkofahi, H., Alawneh, H., Skjellum, A. (2024). MitM attacks on intellectual property and integrity of additive manufacturing systems: A security analysis. Computers & Security, 140, Article 103810. https://doi.org/10.1016/j.cose.2024.103810
Al Mamun, A., Liu, C., Kan, C., Tian, W. (2022). Securing cyber-physical additive manufacturing systems by in-situ process authentication using streamline video analysis. Journal of Manufacturing Systems, 62, 429–440. https://doi.org/10.1016/j.jmsy.2021.12.007
Stouffer, K., Pease, M., Tang, C., Zimmerman, T., Pillitteri, V., Lightman, S., Hahn, A., Saravia, S., Sherule, A., Thompson, M. (2023). Guide to operational technology (OT) security (NIST Special Publication 800-82, Revision 3). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.SP.800-82r3
Oleinikov, I. A., Sribna, I. M. (2025). Development of a methodology for detecting 3D printing defects based on the ResNet model. Telecommunication and Information Technologies, 1, 111–119. https://doi.org/10.31673/2412-4338.2025.014545
Bondarchuk, A. P., Oleinikov, I. A., Bazhan, T. O. (2024). Application of machine learning methods to 3D printer control. Telecommunication and Information Technologies, 1, 4–15. https://doi.org/10.31673/2412-4338.2024.010415
Rybalchenko, L., Prokopovych-Tkachenko, D. (2025). A Bayesian approach to training deep neural networks for risk management in cyber-physical systems. Cybersecurity: Education, Science, Technique, 4(28), 234–245. https://doi.org/10.28925/2663-4023.2025.28.785
Palko, D. (2025). Intelligent models for cybersecurity risk assessment in distributed systems based on a neural network approach. Cybersecurity: Education, Science, Technique, 3(27), 429–448. https://doi.org/10.28925/2663-4023.2025.27.764
Zhdanova, Yu., Shevchenko, S., Spasiteleva, S., Sokulskyi, O. (2024). Decision-making based on linear optimization in information security risk management. Cybersecurity: Education, Science, Technique, 1(25), 330–343. https://doi.org/10.28925/2663-4023.2024.25.330343
Kret, T., Martseniuk, Ye. (2025). An integrated approach to threat modeling in artificial intelligence systems. Cybersecurity: Education, Science, Technique, 2(30), 555–567. https://doi.org/10.28925/2663-4023.2025.30.993
Zhebka, V., & Chytulian, V. (2026). Gaussian Process Regression Model for Performance Prediction of Enterprise Web Systems under Configurable Workloads. Telecommunication and Information Technologies, 2(91), 215–223. https://doi.org/10.31673/2412-4338.2026.029118
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Іван Олейніков, Юрій Базака, Софія Огнева, Сергій Обушний
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
