Pages about teaching are available in German only Zurück zu der Liste der Abschlussarbeiten

Machine Learning Methods for Optimized Industrial Product Acceptance Testing

Title Machine Learning Methods for Optimized Industrial Product Acceptance Testing

Manufacturing processes are highly automated and standardized. However, the newest information technology, such as machine learning, is often not utilized throughout the entire product lifecycle. The active integration of smart and innovative solutions in the manufacturing process, which is a key task in Industry 4.0, results in benefits ranging from increased visibility into operations, to substantial cost savings, to faster production times. Using sensors for data collection and cloud computing for storing and organizing makes data usable in the analysis and implementation of suitable machine learning models. This enables manufacturers to improve production, optimize operations, and gain a better understanding of the product and manufacturing phases. For example, the product test phase and decisions within it could be completely automated by machine learning algorithms to facilitate large production volumes.
This bachelor thesis is written in cooperation with an industry partner producing hydraulic components. Every produced part has to pass a test to ensure compliance with the highest quality standards. A machine learning model that can decide better than simple thresholds on specific aggregations whether a part can or cannot pass a test significantly shortens production time during the testing phase. A reduced production time subsequently results in lower production costs.

Resources BA_MLforProductTesting_OlgaAkulinushkina.pdf (1181 KB)
Thesistype Bachelor Masterthesis
Second Tutor Kotthaus, Helena
Professor Morik, Katharina
Assigned To Akulinushkina, Olga
Status Abgeschlossen
Registered On Sep 3, 2020 9:39:00 AM
Finished On Jan 2, 2021 9:39:00 AM