Hybrid sequence learning with interpretability for multi-class quality prediction in injection molding
Ensuring consistent quality in injection molding remains a critical challenge due to dynamic process variations and the limitations of traditional rule-based inspection methods. This study proposes a novel hybrid deep learning framework that integrates a Transformer encoder with a TabNet classifier...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-09-01
|
| Series: | Results in Engineering |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2590123025024788 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850080336069787648 |
|---|---|
| author | Varathorn Punyangarm Supatchaya Chotayakul |
| author_facet | Varathorn Punyangarm Supatchaya Chotayakul |
| author_sort | Varathorn Punyangarm |
| collection | DOAJ |
| description | Ensuring consistent quality in injection molding remains a critical challenge due to dynamic process variations and the limitations of traditional rule-based inspection methods. This study proposes a novel hybrid deep learning framework that integrates a Transformer encoder with a TabNet classifier to enable interpretable, multi-class defect prediction using time-series part weight data. The Transformer module captures long-range temporal dependencies, while TabNet provides feature-level interpretability through sparse attention masks. The model was trained and validated on real-world data from over 30,000 injection cycles, covering five classes: acceptable part, short shot, flash, sink mark, and warpage. Evaluation results demonstrate that the proposed model significantly outperforms conventional machine learning methods such as Random Forest, XGBoost, CatBoost, and a hybrid deep learning baseline (CNN–TabNet), achieving a macro F1-score of 0.964 and a macro-averaged area under the receiver operating characteristic curve (AUROC) of 0.992. It also maintains high robustness under signal noise and supports inference within 100 milliseconds, enabling near real-time deployment (i.e., high-speed analysis of recent production windows). Importantly, the model offers actionable insights through built-in explainability mechanisms, helping operators understand and trace the root causes of predicted defects. This research contributes a scalable, low-cost, and interpretable solution for proactive quality monitoring, paving the way for practical adoption of explainable AI in smart manufacturing environments. |
| format | Article |
| id | doaj-art-8438982b59c749b3a0ee110e739a8d35 |
| institution | DOAJ |
| issn | 2590-1230 |
| language | English |
| publishDate | 2025-09-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Results in Engineering |
| spelling | doaj-art-8438982b59c749b3a0ee110e739a8d352025-08-20T02:44:56ZengElsevierResults in Engineering2590-12302025-09-012710640810.1016/j.rineng.2025.106408Hybrid sequence learning with interpretability for multi-class quality prediction in injection moldingVarathorn Punyangarm0Supatchaya Chotayakul1Department of Industrial Engineering, Faculty of Engineering, Srinakharinwirot University, Ongkharak Campus, Nakhon Nayok 26120, ThailandCorresponding author.; Department of Industrial Engineering, Faculty of Engineering, Srinakharinwirot University, Ongkharak Campus, Nakhon Nayok 26120, ThailandEnsuring consistent quality in injection molding remains a critical challenge due to dynamic process variations and the limitations of traditional rule-based inspection methods. This study proposes a novel hybrid deep learning framework that integrates a Transformer encoder with a TabNet classifier to enable interpretable, multi-class defect prediction using time-series part weight data. The Transformer module captures long-range temporal dependencies, while TabNet provides feature-level interpretability through sparse attention masks. The model was trained and validated on real-world data from over 30,000 injection cycles, covering five classes: acceptable part, short shot, flash, sink mark, and warpage. Evaluation results demonstrate that the proposed model significantly outperforms conventional machine learning methods such as Random Forest, XGBoost, CatBoost, and a hybrid deep learning baseline (CNN–TabNet), achieving a macro F1-score of 0.964 and a macro-averaged area under the receiver operating characteristic curve (AUROC) of 0.992. It also maintains high robustness under signal noise and supports inference within 100 milliseconds, enabling near real-time deployment (i.e., high-speed analysis of recent production windows). Importantly, the model offers actionable insights through built-in explainability mechanisms, helping operators understand and trace the root causes of predicted defects. This research contributes a scalable, low-cost, and interpretable solution for proactive quality monitoring, paving the way for practical adoption of explainable AI in smart manufacturing environments.http://www.sciencedirect.com/science/article/pii/S2590123025024788Injection moldingDefect predictionTransformerTabnetExplainable AI |
| spellingShingle | Varathorn Punyangarm Supatchaya Chotayakul Hybrid sequence learning with interpretability for multi-class quality prediction in injection molding Results in Engineering Injection molding Defect prediction Transformer Tabnet Explainable AI |
| title | Hybrid sequence learning with interpretability for multi-class quality prediction in injection molding |
| title_full | Hybrid sequence learning with interpretability for multi-class quality prediction in injection molding |
| title_fullStr | Hybrid sequence learning with interpretability for multi-class quality prediction in injection molding |
| title_full_unstemmed | Hybrid sequence learning with interpretability for multi-class quality prediction in injection molding |
| title_short | Hybrid sequence learning with interpretability for multi-class quality prediction in injection molding |
| title_sort | hybrid sequence learning with interpretability for multi class quality prediction in injection molding |
| topic | Injection molding Defect prediction Transformer Tabnet Explainable AI |
| url | http://www.sciencedirect.com/science/article/pii/S2590123025024788 |
| work_keys_str_mv | AT varathornpunyangarm hybridsequencelearningwithinterpretabilityformulticlassqualitypredictionininjectionmolding AT supatchayachotayakul hybridsequencelearningwithinterpretabilityformulticlassqualitypredictionininjectionmolding |