Intelligent Defect Prediction and Yield Optimization in PCB Assembly Lines Using Deep Learning-Based Vision Systems and Reinforcement Learning Process Control
DOI:
https://doi.org/10.15662/IJEETR.2026.0802476Keywords:
PCB defect detection, deep learning, ViT-Mamba, YOLOv11, reinforcement learning, PPO, yield optimization, automated optical inspection, Industry 4.0, smart manufacturingAbstract
The escalating complexity of modern printed circuit board (PCB) assembly lines demands intelligent, real-time quality control systems that transcend traditional automated optical inspection (AOI) capabilities. This paper presents a novel integrated framework that couples deep learning-based vision systems with reinforcement learning (RL) process control to achieve simultaneous defect prediction and manufacturing yield optimization. The vision subsystem employs a hybrid ViT-Mamba architecture augmented with an Efficient Multi-Scale Attention (EMA) module for high-fidelity defect localization, achieving a mean Average Precision (mAP@0.5) of 99.0% across six defect categories at 78 frames per second. The process control subsystem utilizes a Proximal Policy Optimization (PPO) RL agent that dynamically adjusts solder paste volume, reflow oven temperature profiles, component placement offsets, and conveyor speeds in response to real-time defect feedback signals. The closed-loop system converged in approximately 290 training episodes, raising manufacturing yield from a baseline of 87.5% to 97.2%—a relative improvement of 11.1%—while simultaneously reducing the defect rate from 8.7% to 1.9%. Comprehensive ablation experiments confirm that the synergy between vision and control subsystems is essential for peak performance. The proposed architecture represents a significant step toward fully autonomous, zero-defect PCB manufacturing aligned with Industry 4.0 and smart electronics production paradigms.
References
1. Fonseca, L. A., Iano, Y., Oliveira, G. G., Vaz, G. C., Carnielli, G. P., Pereira, J. C., & Arthur, R. (2024). Automatic printed circuit board inspection: A comprehensible survey. Discover Artificial Intelligence, 4(1), 10. https://doi.org/10.1007/s44163-023-00081-5
2. Gao, Y., Zhang, H., et al. (2025). Enhanced YOLOv11 framework for high precision defect detection in printed circuit boards. Scientific Reports, 15, 27415. https://doi.org/10.1038/s41598-025-27415-w
3. Immordino, A., Stöckermann, P., Hayen, N., et al. (2025). Explainable AI for reinforcement learning based dynamic scheduling solutions in semiconductor manufacturing. Journal of Intelligent Manufacturing. https://doi.org/10.1007/s10845-025-02631-3
4. Kieu, X. T., Nguyen, V. T., Chu, D. T., Van, X. H., Van, M., Su, S. F., & Phan, X. T. (2025). Deep learning-enhanced defects detection for printed circuit boards. Results in Engineering, 25, 104067. https://doi.org/10.1016/j.rineng.2025.104067
5. Mangukiya, M., & Miyani, H. (2025). AI-driven process optimization in electronic manufacturing: From PCB assembly to system integration. In 2025 IEEE 5th International Conference on ICT in Business Industry & Government (ICTBIG) (pp. 1–6). IEEE. https://doi.org/10.1109/ICTBIG68706.2025.11323740
6. Nia, A., Shen, W., & Vogel-Heuser, B. (2025). Deep reinforcement learning for optimal planning of production line maintenance with deterioration. Reliability Engineering & System Safety, 262, 111767. https://doi.org/10.1016/j.ress.2025.111767
7. Niaz, A., Umraiz, M., & Soomro, S. (2025). Vision transformer and Mamba-attention fusion for high-precision PCB defect detection. PLOS ONE, 20(9), e0331175. https://doi.org/10.1371/journal.pone.0331175
8. Ong, T. Y., et al. (2025). Review of solder joint vision inspection for industrial applications. International Journal of Advanced Manufacturing Technology. https://doi.org/10.1007/s00170-025-15383-4
9. Peres, R. S., et al. (2024). Reinforcement learning for autonomous process control in Industry 4.0: Advantages and challenges. Applied Artificial Intelligence. https://doi.org/10.1080/08839514.2024.2383101
10. Sun, X., Shen, W., Fan, J., Vogel-Heuser, B., Bi, F., & Zhang, C. (2025). Deep reinforcement learning-based multi-objective scheduling for distributed heterogeneous hybrid flow shops with blocking constraints. Engineering, 46(3), 293–306. https://doi.org/10.1016/j.eng.2024.11.033
11. Villanueva, A., et al. (2025). Deep learning-based solder joint defect detector. International Journal of Advanced Manufacturing Technology. https://doi.org/10.1007/s00170-025-15460-8
12. Jothilingam, P. (2020). AI-enabled predictive maintenance for optimizing plant operations: Data-driven approaches for fault detection, diagnostics, and lifecycle management. International Journal of Open Publication and Exploration (IJOPE), 8(20), 8. https://ijope.com/index.php/home/article/view/211/189
13. Wang, Y., Wu, B., Zhang, L., et al. (2025). Enhanced PCB defect detection via HSA-RTDETR on RT-DETR. Scientific Reports, 15, 31783. https://doi.org/10.1038/s41598-025-11394-z
14. Xie, J., Guo, Y., Liu, D., et al. (2025). A multimodal fusion method for soldering quality online inspection. Journal of Intelligent Manufacturing, 36, 3271–3284. https://doi.org/10.1007/s10845-024-02413-3
15. Yang, W., et al. (2025). Automated detection and classification of soldering defects in printed circuit boards using deep learning and optical and thermal imaging. Journal of Intelligent Manufacturing. https://doi.org/10.1007/s10845-025-02691-5
16. Zhou, X., et al. (2025). A comprehensive review of research on surface defect detection of PCBs based on machine vision. Advanced Engineering Informatics, 68, Article 102500. https://doi.org/10.1016/j.aei.2025.102500
