 |
International Society of Science and Applied Technologies |
| Smart Manufacturing Acceptance Rate Prediction Model: A Deep Learning Approach | ||||
| Author | Chiao-Tzu Huang
|
|||
| Co-Author(s) |
Yi Peng Liao; Cheng-Kai Hung
|
|||
| Abstract | This paper presents a novel Smart Acceptance Prediction Model that transforms In-Process Quality Control (IPQC) by predicting customer acceptance rates at early production stages. By implementing deep learning techniques—specifically a Deep GAN LRCN-GPR architecture with Model Adaptation (Transfer Learning with Domain Adaptation) framework—our model identifies potential defects at upstream workstations, preventing defective products from flowing into downstream processes. This enables intelligent production line segregation, where only output from problematic workstations undergoes 100% inspection while maintaining high precision in quality prediction. The proposed model demonstrates remarkable performance with up to 95% prediction accuracy across diverse production scenarios. The implementation of this Smart Acceptance Prediction Model delivers a crucial benefit by significantly reducing the burden of 100% inspection on production lines. By achieving consistently high prediction accuracy, the model enables manufacturers to confidently transition from comprehensive inspection of all products to a more efficient sampling inspection methodology. This shift not only alleviates production bottlenecks and resource constraints but also maintains rigorous quality standards while improving overall production throughput and operational efficiency. The model thus provides a data-driven foundation for optimizing inspection protocols without compromising product quality or customer satisfaction
|
|||
| Keywords | Smart Acceptance Prediction, Quality Control, Transfer Learning, Model Adaptation | |||
| Article #: RQD2025-111 | ||||
Proceedings of 30th ISSAT International Conference on Reliability & Quality in Design |