Lightweight You Only Look Once-based automatic defect detection in wire arc additive manufacturing

Abstract

Wire arc additive manufacturing (WAAM) technology has achieved significant advancements in fabricating complex metal components, yet defects such as surface porosity, lack of fusion, and slag inclusion continue to compromise quality and efficiency. To address challenges in detecting welding slag and micro-pores, we introduce an enhanced You Only Look Once (YOLO)v8n-Attention-Refined Feature Module (ARFM) architecture integrating receptive field attention convolution, an attention-based feature pyramid network, and the Focaler-minimum point distance intersection over union metric. This integration markedly improves the precision and resilience of defect recognition. Moreover, we assessed the defect detection capabilities of YOLOv3t, YOLOv5n, YOLOv6n, YOLOv8n, YOLOv9t, and YOLOv10n in WAAM and developed a composite defect dataset incorporating aluminum, titanium, and stainless steel specimens. The experiments reveal that YOLOv8n achieved the highest overall detection effectiveness (mean average precision [mAP]@0.5 = 0.921). YOLOv10n exhibited marginal superiority in slag inclusion detection (mAP@0.5 = 0.886) with a peak throughput of 54.35 frames per second (FPS), while YOLOv5n converged the fastest within 500 epochs. Following the introduction of ARFM, mAP@0.5 increased to 0.944 (+2.4%), and slag inclusion detection reached 0.929 (+7.3%). However, the frame rate declined to 38 FPS, which suffices for basic real-time monitoring but remains inadequate for high real-time scenarios. This study provides an important reference for deploying YOLO-series models for real-time monitoring of WAAM and closed-loop control systems.

Keywords

Wire arc additive manufacturing / Lightweight You Only Look Once / Defect detection / Asymptotic feature pyramid network / Receptive field augmented convolution / Focaler-minimum point distance intersection over union

Funding

This work is sponsored by the CNPC Innovation Foundation (Grant No. 2024DQ02-0306), the Basic Business Research Fees of Central Universities (Grant No. 22CX06049A), the Natural Science Foundation of Qingdao (Grant No. 23-2-1-83-zyyd-jch), and the Natural Science Foundation of Shandong Province (Grant No. ZR2023QE164).

Conflict of Interest

The authors declare that they have no competing interests.

References

[1] Li R, Wang G, Ding Y, et al. Optimization of the geometry for the end lateral extension path strategy to fabricate intersections using laser and cold metal transfer hybrid additive manufacturing. Addit Manuf. 2020;36:101546. doi: 10.1016/j.addma.2020.101546

[2] Bian Y, Hu T, Lv Z, et al. Bone tissue engineering for treating osteonecrosis of the femoral head. Exploration. 2023;3(2):20210105. doi: 10.1002/EXP.20210105

[3] Mahmud MAP, Tat T, Xiao X, Adhikary P, Chen J. Advances in 4D-printed physiological monitoring sensors. Exploration. 2021;1(3):20210033. doi: 10.1002/exp.20210033

[4] Wang G, Mao X, Wang W, Wang X, Li S, Wang Z. Bioprinted research models of urological malignancy. Exploration. 2024;4(4):20230126. doi: 10.1002/exp.20230126

[5] Sun C, Wang Y, McMurtrey MD, Jerred ND, Liou F, Li J. Additive manufacturing for energy: A review. Appl Energy. 2021;282:116041. doi: 10.1016/j.apenergy.2020.116041

[6] Li R, Ma H, Zeng K, et al. Differential evolution-optimized multi-output support vector regression-based prediction of weld bead morphology in wire-fed laser-arc directed energy deposition of 2319 aluminum alloy. Addit Manuf Front. 2025;4:200203. doi: 10.1016/j.amf.2025.200203

[7] Li JN, Gao D, Lu Y, Hao ZP, Wang ZQ. Mechanical properties and microstructure evolution of additive manufactured 316L stainless steel under dynamic loading. Mater Sci Eng A. 2022;855:143896. doi: 10.1016/j.msea.2022.143896

[8] Li R, Wang G, Zhao X, et al. Effect of path strategy on residual stress and distortion in laser and cold metal transfer hybrid additive manufacturing. Addit Manuf. 2021;46:102203. doi: 10.1016/j.addma.2021.102203

[9] Tomar B, Shiva S, Nath T. A review on wire arc additive manufacturing: Processing parameters, defects, quality improvement and recent advances. Mater Today Commun. 2022;31:103739. doi: 10.1016/j.mtcomm.2022.103739

[10] Wei X. Additive manufacturing as a tool for high-throughput experimentation. J Mater Inf. 2022;2(3):12. doi: 10.20517/jmi.2022.19

[11] Lin X, Zhu K, Fuh JYH, Duan X. Metal-based additive manufacturing condition monitoring methods: From measurement to control. ISA Trans. 2022;120:147-166. doi: 10.1016/j.isatra.2021.03.001

[12] Du Plessis A, Macdonald E. Hot isostatic pressing in metal additive manufacturing: X-ray tomography reveals details of pore closure. Addit Manuf. 2020;34:101191. doi: 10.1016/j.addma.2020.101191

[13] Honarvar F, Varvani-Farahani A. A review of ultrasonic testing applications in additive manufacturing: Defect evaluation, material characterization, and process control. Ultrasonics. 2020;108:106227. doi: 10.1016/j.ultras.2020.106227

[14] Erik W. Porosity and density measurement of additively manufactured components: A comparative analysis of measurement methods across processes and materials. MSAM. 2025;4(2):025090010. doi: 10.36922/MSAM025090010

[15] Benbarrad T, Salhaoui M, Kenitar SB, Arioua M. Intelligent machine vision model for defective product inspection based on machine learning. J Sensor Actuator Netw. 2021;10(1):7. doi: 10.3390/jsan10010007

[16] Sani AR, Zolfagharian A, Kouzani AZ. Artificial intelligence‐augmented additive manufacturing: Insights on closed‐loop 3D printing. Adv Intell Syst. 2024;6(10):2400102. doi: 10.1002/aisy.202400102

[17] Suyama FM, Delgado MR, Dutra da Silva R, Centeno TM. Deep neural networks based approach for welded joint detection of oil pipelines in radiographic images with Double Wall Double Image exposure. NDT E Int. 2019;105:46-55. doi: 10.1016/j.ndteint.2019.05.002

[18] Xia C, Pan Z, Li Y, Chen J, Li H. Vision-based melt pool monitoring for wire-arc additive manufacturing using deep learning method. Int J Adv Manuf Technol. 2022;120(1):551-562. doi: 10.1007/s00170-022-08811-2

[19] Li R, Ma H, Wang R, et al. Application of unsupervised learning methods based on video data for real-time anomaly detection in wire arc additive manufacturing. J Manuf Processes. 2025;143:37-55. doi: 10.1016/j.jmapro.2025.03.113

[20] Abranovic B, Sarkar S, Chang-Davidson E, Beuth J. Melt pool level flaw detection in laser hot wire directed energy deposition using a convolutional long short-term memory autoencoder. Addit Manuf. 2024;79:103843. doi: 10.1016/j.addma.2023.103843

[21] Li W, Zhang H, Wang G, et al. Deep learning based online metallic surface defect detection method for wire and arc additive manufacturing. Robot Comput Integr Manuf. 2023;80:102470. doi: 10.1016/j.rcim.2022.102470

[22] Zhou X, Zheng S, Li R, et al. Improved YOLOv5-based pore defect detection algorithm for wire arc additive manufacturing. Mater Today Commun. 2024;39:108710. doi: 10.1016/j.mtcomm.2024.108710

[23] Sani AR, Zolfagharian A, Kouzani AZ. Automated defects detection in extrusion 3D printing using YOLO models. J Intell Manuf. 2024;1-21. doi: 10.1007/s10845-024-02543-8

[24] Chandra M, Rajak S, Vimal KEK. Deep learning-based framework for the observation of real-time melt pool and detection of anomaly in wire-arc additive manufacturing. Mater Manuf Processes. 2024;39(6):761-777. doi: 10.1080/10426914.2023.2254386

[25] Huang Y, Zhou X, Xiong X, Fu Y. Efficient defect detection method for wire and arc additive manufacturing based on modified YOLOv8 model. J Nondestruct Eval. 2025;44(2):47. doi: 10.1007/s10921-025-01181-1

[26] Sohan M, Sai Ram T, Rami Reddy CV. A Review on YOLOv8 and Its Advancements. Singapore: Springer Nature; 2024. p. 529-545. doi: 10.1007/978-981-99-7962-2_39

[27] Wang G, Chen Y, An P, Hong H, Hu J, Huang T. UAV-YOLOv8: A small-object-detection model based on improved YOLOv8 for UAV aerial photography scenarios. Sensors. 2023;23(16):7190. doi: 10.3390/s23167190

[28] Yang G, Lei J, Zhu Z, Cheng S, Feng Z, Liang R. AFPN: Asymptotic feature pyramid network for object detection. In: 2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC); 2023. p. 2184-2189. doi: 10.1109/SMC53992.2023.10394415

[29] Zhang X, Liu C, Yang D, et al. RFAConv: Innovating Spatial Attention and Standard Convolutional Operation. [arXiv Preprint]; 2023.

[30] Wu J, Zhang J, Zhu J, et al. Multi-scale convolution and dynamic task interaction detection head for efficient lightweight plum detection. Food Bioprod Processing. 2025;149:353-367. doi: 10.1016/j.fbp.2024.12.007

[31] Zhang H, Spa Z. Focaler-IoU: More Focused Intersection over Union Loss. [arXiv Preprint]; 2024.

[32] Siliang M, Xpa Y. Mpdiou: A Loss for Efficient and Accurate Bounding Box Regression [arXiv Preprint]; 2023.

[33] Guo Y, Chen S, Zhan R, Wang W, Zhang J. LMSD-YOLO: A lightweight YOLO algorithm for multi-scale SAR ship detection. Remote Sens. 2022;14(19):4801. doi: 10.3390/rs14194801

[34] Liao Y, Li L, Xiao H, Xu F, Shan B, Yin H. YOLO-MECD: Citrus detection algorithm based on YOLOv11. Agronomy. 2025;15(3):687. doi: 10.3390/agronomy15030687

[35] Lu J, Yu R, Zhang H, et al. Highly reflective metal industrial parts surface detection with polarization vision and modified YOLO. Digit Signal Processing. 2025;163:105238. doi: 10.1016/j.dsp.2025.105238

[36] Chen J, Mai H, Luo L, Chen X, Wu K. Effective Feature Fusion Network in BIFPN for Small Object Detection. United States: IEEE; 2021. p. 699-703. doi: 10.1109/icip42928.2021.9506347

[37] Zhang Y, Jiang X, Qiu B, et al. GFPN-YOLOv8m: A Deep Learning Network for Astronomical Source Detection. United States: IEEE; 2024. p. 473-477. doi: 10.1109/ntci64025.2024.10776286

[38] Chen Y, Dai X, Liu M, Chen D, Yuan L, Liu Z. Dynamic Convolution: Attention Over Convolution Kernels. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR); 2020. p. 11030-11039.

[39] Li J, Wen Y, He L. Scconv: Spatial and Channel Reconstruction Convolution for Feature Redundancy. United States: IEEE; 2023. p. 6153-6162.

[40] Rezatofighi H, Tsoi N, Gwak J, Sadeghian A, Reid I, Savarese S. Generalized Intersection Over Union: A Metric and a Loss for Bounding Box Regression. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR); 2019. p. 658-666.

[41] Gevorgyan Z. SIoU Loss: More Powerful Learning for Bounding Box Regression. [arXiv Preprint]; 2022.

[42] Dai J, Qi H, Xiong Y, et al. Deformable convolutional networks. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV); 2017. p. 764-773.

[43] Kervadec H, Bouchtiba J, Desrosiers C, Granger E, Dolz J, Ayed IB. Boundary Loss for Highly Unbalanced Segmentation. Presented at: Proceedings of The 2^nd International Conference on Medical Imaging with Deep Learning. Proceedings of Machine Learning Research; 2019. Available from: https://proceedings.mlr.press/v102/ kervadec19a.html [Last accessed on 2025 Jun 20].