Bretones Cassoli Nevejan, Beatriz (2024)
Graph representation learning for predictive quality in multi-stage discrete manufacturing.
Technische Universität Darmstadt
doi: 10.26083/tuprints-00027808
Ph.D. Thesis, Primary publication, Publisher's Version
Text
Dissertation_BBCN.pdf Copyright Information: CC BY-SA 4.0 International - Creative Commons, Attribution ShareAlike. Download (6MB) |
Item Type: | Ph.D. Thesis | ||||
---|---|---|---|---|---|
Type of entry: | Primary publication | ||||
Title: | Graph representation learning for predictive quality in multi-stage discrete manufacturing | ||||
Language: | English | ||||
Referees: | Metternich, Prof. Dr. Joachim ; Groche, Prof. Dr. Peter | ||||
Date: | 31 July 2024 | ||||
Place of Publication: | Darmstadt | ||||
Collation: | xxii, 134 Seiten | ||||
Date of oral examination: | 2 July 2024 | ||||
DOI: | 10.26083/tuprints-00027808 | ||||
Abstract: | This thesis addresses the need for predictive quality control methods in multi-stage discrete manufacturing processes to enhance operational efficiency and product quality. Traditional end-of-line quality control practices are being replaced or supplemented by data-driven approaches to mitigate manual testing dependencies. Leveraging graph representation learning, inspired by successful applications in biochemistry, this research proposes a novel approach for predictive quality control, aiming to capture complex interdependencies within manufacturing processes. The primary objective is to evaluate the efficacy of this approach compared to alternative machine learning methods. Through comprehensive theoretical and empirical analysis, this thesis makes two significant contributions. Firstly, it introduces a novel graph representation learning approach tailored for manufacturing data modeling and product quality classification. This approach offers a systematic guideline for implementation across diverse manufacturing contexts. Secondly, empirical validation through two distinct case studies demonstrates the superior performance of the proposed method over conventional machine learning techniques. The results support its potential for enhancing product quality classification and streamlining quality control operations in multi-stage discrete manufacturing. Overall, this research contributes to advancing predictive quality control methodologies in multi-stage discrete manufacturing processes, offering practical insights and guidelines for industry adoption in pursuit of enhanced operational efficiency and product quality. |
||||
Alternative Abstract: |
|
||||
Status: | Publisher's Version | ||||
URN: | urn:nbn:de:tuda-tuprints-278088 | ||||
Classification DDC: | 600 Technology, medicine, applied sciences > 620 Engineering and machine engineering | ||||
Divisions: | 16 Department of Mechanical Engineering > Institute of Production Technology and Machine Tools (PTW) 16 Department of Mechanical Engineering > Institute of Production Technology and Machine Tools (PTW) > Management of Industrial Production |
||||
Date Deposited: | 31 Jul 2024 12:03 | ||||
Last Modified: | 01 Aug 2024 07:36 | ||||
URI: | https://tuprints.ulb.tu-darmstadt.de/id/eprint/27808 | ||||
PPN: | 520245504 | ||||
Export: |
View Item |