IMPROVEMENT OF INFORMATION AND MEASURING TECHNOLOGIES IN WELDING AND INSTALLATION OF METAL STRUCTURES
DOI:
https://doi.org/10.35546/kntu2078-4481.2025.2.2.2Keywords:
welding, information and measurement technologies, artificial intelligence, machine learning, feedback, quality control, adaptive systems, CNN, XGBoostAbstract
This article considers an approach to improving information and measurement technologies during welding and installation of metal structures, as well as presents the results of development and testing of the proposed integrated system. Particular attention is paid to the development and application of artificial intelligence algorithms, such as machine learning and deep neural networks, to improve the accuracy, speed, and adaptability of information and measurement technologies. Their role in detecting deviations, predicting the development of defects and automatic adaptation of welding parameters to changes in the production environment is considered. The proposed integrated system combines a sensor subsystem, a data preprocessing module for cleaning and structuring information, a neural network analytical unit (CNN for visual control and XGBoost for parameter classification), as well as a feedback loop that provides adaptive adjustment of parameters in real time and prompt response to deviations. Thanks to the implementation of a full closed cycle – from data collection, pre-processing and analysis to automatic parameter adjustment – the impact of the human factor is minimized, process reliability is increased and the number of defects is significantly reduced. Such a cycle ensures continuous updating of data, their thorough checking, quick response and flexible adjustment of welding parameters in accordance with changes in the production environment. The use of the Python ecosystem (Pandas, NumPy, TensorFlow, Scikit-learn) for data analysis, the use of sensors for reading key parameters and adaptive control via the MQTT and PLC protocols ensure a 27 % reduction in defects and process stability. The presented approach combines modern machine learning methods with a complex IoT architecture in detail, emphasizing their interaction and opening up prospects for the further development of intelligent production systems that are capable of self-learning and adaptation in real time.
References
Roosvel Soto-Diaz, Mauricio Vásquez-Carbonell, Jose Escorcia-Gutierrez, A review of artificial intelligence techniques for optimizing friction stir welding processes and predicting mechanical properties, Engineering Science and Technology, an International Journal, Volume 62, 2025, 101949, ISSN 2215-0986, https://doi.org/10.1016/ j.jestch.2025.101949
Jalal Taheri Kahnamouei, Mehrdad Moallem, Advancements in control systems and integration of artificial intelligence in welding robots: A review, Ocean Engineering, Volume 312, Part 3, 2024, 119294, ISSN 0029-8018. https://doi.org/10.1016/j.oceaneng.2024.119294
Michael Luttmer, Matthias Weigold, Heiko Thaler, Jürgen Dongus, Anton Hopf, Towards data-driven quality monitoring for advanced metal inert gas welding processes in body-in-white, Journal of Manufacturing Systems, Volume 77, 2024, Pages 875–891, ISSN 0278-6125, https://doi.org/10.1016/j.jmsy.2024.10.013
Mustapha Belmouadden, Camélia Dadouchi, Robert Pellerin, Artificial intelligence applied in adaptive manufacturing process monitoring: a state-of-the-art in the era of automation., Procedia Computer Science, Volume 256, 2025, Pages 47–54, ISSN 1877-0509, https://doi.org/10.1016/j.procs.2025.02.094
Nijat Mehdiyev, Maxim Majlatow, Peter Fettke, Integrating permutation feature importance with conformal prediction for robust Explainable Artificial Intelligence in predictive process monitoring, Engineering Applications of Artificial Intelligence, Volume 149, 2025, 110363, ISSN 0952-1976, https://doi.org/10.1016/j.engappai.2025.110363
D. Görick, L. Larsen, M. Engelschall, A. Schuster, Quality Prediction of Continuous Ultrasonic Welded Seams of High-Performance Thermoplastic Composites by means of Artificial Intelligence, Procedia Manufacturing, Volume 55, 2021, Pages 116–123, ISSN 2351-9789, https://doi.org/10.1016/j.promfg.2021.10.017
K. Sabatakakis, N. Bourlesas, H. Bikas, A. Papacharalampopoulos, P. Stavropoulos, Laser Welding of dissimilar cell tabs: Extracting physics semantics from infrared (IR) emissions as process monitoring data, Procedia CIRP, Volume 121, 2024, Pages 222-227, ISSN 2212-8271, https://doi.org/10.1016/j.procir.2023.09.251
Henrique H. L. Núñez, Li-Wei Hsu, Kandice S. B. Ribeiro, Antti Salminen, Wallace M. Bessa, In-situ monitoring and online prediction of keyhole depth in laser welding by coaxial imaging, Procedia CIRP, Volume 124, 2024, Pages 793–796, ISSN 2212-8271, https://doi.org/10.1016/j.procir.2024.08.227
Hyungi Byun, Han Gil Lee, Beom Kyu Kim, Geun Dong Song, Bongsoo Lee, Defect monitoring system of the internal structures of a sodium fast reactor using an artificial intelligence model, Nuclear Engineering and Technology, Volume 56, Issue 12, 2024, Pages 5405–5413, ISSN 1738-5733, https://doi.org/10.1016/j.net.2024.07.049
Bin Shen, Jun Lu, Yiming Wang, Dongli Chen, Jing Han, Yi Zhang, Zhuang Zhao, Multimodal-based weld reinforcement monitoring system for wire arc additive manufacturing, Journal of Materials Research and Technology, Volume 20, 2022, Pages 561–571, ISSN 2238-7854, https://doi.org/10.1016/j.jmrt.2022.07.086
Mobina Mobaraki, Soodeh Ahani, Ringo Gonzalez, Kwang Moo Yi, Klaske Van Heusden, Guy A. Dumont, Vision-based seam tracking for GMAW fillet welding based on keypoint detection deep learning model, Journal of Manufacturing Processes, Volume 117, 2024, Pages 315–328, ISSN 1526-6125, https://doi.org/10.1016/j.jmapro.2024.03.006
Tim Raffin, Andreas Mayr, Marcel Baader, Nadine Laube, Alexander Kühl, Jörg Franke, Potentials of few-shot learning for quality monitoring in laser welding of hairpin windings, Procedia CIRP, Volume 118, 2023, Pages 901–906, ISSN 2212-8271, https://doi.org/10.1016/j.procir.2023.06.155
Nik Weisbrod, Joachim Metternich, Application of a concept for ML-driven closed-loop quality control in laser beam welding, Procedia CIRP, Volume 126, 2024, Pages 739–744, ISSN 2212-8271, https://doi.org/10.1016/j.procir.2024.08.301
Nuttapong Chuenmee, Nattachai Phothi, Kontorn Chamniprasart, Sorada Khaengkarn, Jiraphon Srisertpol, Machine learning for predicting resistance spot weld quality in automotive manufacturing, Results in Engineering, Volume 25, 2025, 103570, ISSN 2590-1230, https://doi.org/10.1016/j.rineng.2024.103570
J. J. Valdiande, M. Martínez-Minchero, A. Cobo, J. M. Lopez-Higuera, J. Mirapeix, On-line monitoring and defect detection of arc-welding via plasma optical spectroscopy and LIBS, Spectrochimica Acta Part B: Atomic Spectroscopy, Volume 194, 2022, 106474, ISSN 0584–8547, https://doi.org/10.1016/j.sab.2022.106474
David Curiel, Alfredo Suárez, Fernando Veiga, Eider Aldalur, Pedro Villanueva, Advanced welding automation: Intelligent systems for multipass welding in Butt Double V-Groove and Tee Double Bevel configurations, MethodsX, Volume 13, 2024, 103027, ISSN 2215-0161, https://doi.org/10.1016/j.mex.2024.103027
Huangyi Qu, Yi Cai, Improved semantic segmentation method for weld penetration prediction of TIG welding with dual ellipsoid heat source, Manufacturing Letters, Volume 41, Supplement, 2024, Pages 310–319, ISSN 2213-8463, https://doi.org/10.1016/j.mfglet.2024.09.037
