Ontology-Enhanced Machine Learning: A Bosch Use Case of Welding Quality Monitoring

• DOI: 10.1007/978-3-030-62466-8_33
• Publication Date: 2020-01-01
• Authors: Yulia Svetashova; Baifan Zhou; Tim Pychynski; Stefan Schmidt; York Sure-Vetter; Ralf Mikut; Evgeny Kharlamov
• Keywords: Design for Manufacture; Ontology; Collaborative Manufacturing

In the automotive industry, welding is a critical process of automated manufacturing and its quality monitoring is important. IoT technologies behind automated factories enable adoption of Machine Learning (ML) approaches for quality monitoring. Development of such ML models requires collaborative work of experts from different areas, including data scientists, engineers, process experts, and managers. The asymmetry of their backgrounds, the high variety and diversity of data relevant for quality monitoring pose significant challenges for ML modeling. In this work, we address these challenges by empowering ML-based quality monitoring methods with semantic technologies. We propose a system, called SemML, for ontology-enhanced ML pipeline development. It has several novel components and relies on ontologies and ontology templates for task negotiation and for data and ML feature annotation. We evaluated SemML on the Bosch use-case of electric resistance welding with very promising results.