Machine Learning-Driven Automated Quality Inspection of Steel Plates in Industrial Production Lines
DOI:
https://doi.org/10.63282/3117-5481/AIJCST-V8I3P105Keywords:
Artificial Intelligence, Deep Learning, Quality Control, Visual Inspection, Smart Manufacturing, Defect DetectionAbstract
Steel is a vital commodity in many different industries and the building trades. It's crucial to have high-quality steel plates to avoid any potential problems. An essential step in detecting and classifying surface defects in rolled metal is the pre-distribution quality evaluation. This study provides a CNN+BiLSTM hybrid neural network system that can automatically detect and categorize surface flaws on steel plates. The Faulty Steel Plates dataset forms basis of the technique proposed in this paper. Cleaning dataset, normalizing it using Min-Max, balancing it using SMOTE, and stratifying it are all data preprocessing procedures that are utilized to improve the dataset's quality and classification performance. In order to improve learning, CNN component records most crucial spatial features and the BiLSTM component records the sequential dependency. The performance of the proposed model is compared with traditional machine learning models such as Decision Tree (DT), Random Forest (RF), and K-Nearest Neighbor (KNN). Experimental findings showed that the suggested model (CNN+BiLSTM) outperformed existing models with an F1-score (F1), recall (rec), accuracy (acc), and precision (prec) of 99.7%. The findings prove that the suggested technique is an effective and dependable way to automatically detect surface defects in steel plates in an industrial production environment.
References
[1] A. Krommuang and O. Suwunnamek, “Internet of Things (IoT) Application for Management in Automotive Parts Manufacturing,” Int. J. Adv. Comput. Sci. Appl., vol. 13, no. 4, 2022, doi: 10.14569/IJACSA.2022.0130474.
[2] Y. A. Purmala, “Implementation of machine learning to increase productivity in the manufacturing industry: a literature review.,” Oper. Excell. J. Appl. Ind. Eng., vol. 13, no. 2, p. 267, Jul. 2021, doi: 10.22441/oe.2021.v13.i2.026.
[3] T. Kulkarni, B. Toksha, S. Shirsath, S. Pankade, and A. T. Autee, “Construction and Praxis of Six Sigma DMAIC for Bearing Manufacturing Process,” Mater. Today Proc., vol. 72, pp. 1426–1433, 2023, doi: 10.1016/j.matpr.2022.09.342.
[4] D. S. Das, V. Challagulla, D. P. Balaramesh, D. A. M. A. Mohan, B. V. Jyothi, and D. S. Saravanan, “Deep Learning for Corrosion Monitoring Virtual Sensor and Predictive Modelling Approaches in Industrial Water Pipeline,” The Bioscan, vol. 21, no. 1, pp. 35–45, Jan. 2026, doi: 10.63001/tbs.2026.v21.i01.pp35-45.
[5] T. H. Febriana and H. Hasbullah, “Analysis and Defect Improvement Using FTA, FMEA, and MLR Through DMAIC Phase: Case Study in Mixing Process Tire Manufacturing Industry,” J. Eur. des Systèmes Autom., vol. 54, no. 5, pp. 721–731, Oct. 2021, doi: 10.18280/jesa.540507.
[6] S. Farahmand‐Tabar and P. Ashtari, “Simultaneous size and topology optimization of 3D outrigger‐braced tall buildings with inclined belt truss using genetic algorithm,” Struct. Des. Tall Spec. Build., vol. 29, no. 13, Sep. 2020, doi: 10.1002/tal.1776.
[7] P. Ashtari, R. Karami, and S. Farahmand-Tabar, “Optimum geometrical pattern and design of real-size diagrid structures using accelerated fuzzy-genetic algorithm with bilinear membership function,” Appl. Soft Comput., vol. 110, p. 107646, Oct. 2021, doi: 10.1016/j.asoc.2021.107646.
[8] S. Farahmand-Tabar and M. Babaei, “Memory-assisted adaptive multi-verse optimizer and its application in structural shape and size optimization,” Soft Comput., vol. 27, no. 16, pp. 11505–11527, Aug. 2023, doi: 10.1007/s00500-023-08349-9.
[9] S. Kilaru, “Automated ETL Intelligence: Metadata-Orchestrated Framework with Rule-Based Heuristics for Monitoring and Reporting,” Int. J. Inf. Electron. Eng., vol. 3, no. 6, p. 14, Aug. 2014, doi: 10.48047/ijiee.2013.3.6.9.
[10] N.-Z. Chen, Z. Zhao, and L. Lin, “A hybrid deep learning method for AE source localization for heterostructure of wind turbine blades,” Mar. Struct., vol. 94, p. 103562, Mar. 2024, doi: 10.1016/j.marstruc.2023.103562.
[11] A. K. Padhy, N. Seshagiri, V. Soni, A. K. Elengovan, G. Babu Thokala, and M. Kumari, “AI-Enabled Autonomous Data Preprocessing: A Scalable Architecture for Intelligent Machine Learning Pipeline Management,” in 2026 6th International Conference on Image Processing and Capsule Networks (ICIPCN), IEEE, Jan. 2026, pp. 1318–1323. doi: 10.1109/ICIPCN67432.2026.11438503.
[12] Y. Liu, K. Xu, and J. Xu, “Periodic Surface Defect Detection in Steel Plates Based on Deep Learning,” Appl. Sci., vol. 9, no. 15, p. 3127, Aug. 2019, doi: 10.3390/app9153127.
[13] R. Tang and K. Mao, “An Improved GANs Model for Steel Plate Defect Detection,” IOP Conf. Ser. Mater. Sci. Eng., vol. 790, no. 1, p. 012110, Mar. 2020, doi: 10.1088/1757-899X/790/1/012110.
[14] D. Soukup and R. Huber-Mörk, “Convolutional Neural Networks for Steel Surface Defect Detection from Photometric Stereo Images,” 2014, pp. 668–677. doi: 10.1007/978-3-319-14249-4_64.
[15] Y. Zhu, Y. Liu, and H. Li, “An Improved RT-DETR Method for Surface Defect Detection of Small Targets on Steel Plates,” in Proceedings of the 2026 International Conference on Artificial Intelligence and Control, New York, NY, USA, NY, USA: ACM, Feb. 2026, pp. 128–134. doi: 10.1145/3807246.3807268.
[16] B. Jiang, P. Zhou, X. Sun, and T. Chai, “Intelligent recognition of steel plate surface defect based on deep convolutional GAN,” Neural Comput. Appl., vol. 37, no. 14, pp. 8331–8345, May 2025, doi: 10.1007/s00521-025-11007-w.
[17] R. Xue et al., “Research on Online Detection Method for Lateral Deviation of Steel Strip in Steel Production Line Based on Machine Vision,” in 2025 IEEE International Conference on Pattern Recognition, Machine Vision and Artificial Intelligence (PRMVAI), IEEE, Jun. 2025, pp. 1–6. doi: 10.1109/PRMVAI65741.2025.11108523.
[18] V. Vasan, N. V. Sridharan, S. Vaithiyanathan, and M. Aghaei, “Detection and classification of surface defects on hot-rolled steel using vision transformers,” Heliyon, vol. 10, no. 19, p. e38498, Oct. 2024, doi: 10.1016/j.heliyon.2024.e38498.
[19] A. Feyzioğlu and Y. S. Taspınar, “Detection of Defects in Rolled Stainless Steel Plates by Machine Learning Models,” Int. J. Appl. Math. Electron. Comput., vol. 11, no. 1, pp. 37–43, Mar. 2023, doi: 10.18100/ijamec.1253191.
[20] X. Feng, X. Gao, and L. Luo, “Research on the application of deep learning based surface defect detection and treatment method for hot rolled strip steel,” in 2022 41st Chinese Control Conference (CCC), IEEE, Jul. 2022, pp. 7361–7366. doi: 10.23919/CCC55666.2022.9902418.
[21] C. Crawford, “Faulty Steel Plates,” Kaggle. [Online]. Available: https://www.kaggle.com/datasets/uciml/faulty-steel-plates
[22] H. Vaidya, K. V. Prasad, R. S, K. S. K, and S. Y, “Enhancing Steel Manufacturing Quality Control: A Deep Learning Approach for Rapid and Accurate Surface Defect Detection Using Pre-Trained Convolutional Neural Network Features and Supervised Classifiers,” Int. Acad. Phys. Sci., vol. 28, no. 1, pp. 1–18, 2024, doi: 1061294/jiaps2024.2811.
[23] B. Tasar, “Comparison Analysis of Machine Learning Algorithms for Steel Plate Fault Detection,” Düzce Univ. J. Sci. Technol., vol. 10, no. 3, pp. 1578–1588, Jul. 2022, doi: 10.29130/dubited.1058467.
