They may not have predicted the outcome of the Euro 2024 soccer final, but newly published work from researchers at the Insight SFI Research Centre for Data Analytics at the University of Galway and Dublin City University, in collaboration with Bosch, is being used to predict missing links between data in football production lines, to help with manufacturing system reliability and overall football product quality.

Manufacturers are integrating various technologies and components into their production lines to make them more efficient and improve throughput. These technologies include robots, sensors, networks, and more, in what is called Industry 4.0. In a football production line, there are various stages involved in manufacturing a football such as cutting, printing, spraying, glueing, moulding/shaping and heating, and components including temperature, speed and pressure sensors and associated electronic/computerised monitoring systems. These advanced production lines are producing huge amounts of data from all of these components and manufacturing stages that has to be integrated. However, when combining this data, sometimes useful connections or links between the data are missing and need to be found.

 By predicting these missing links, using machine learning on what is called Knowledge Graph (KG) technology, this work from the Insight Centre can help to recreate data that can benefit a variety of manufacturing applications, including:

• predictive maintenance (using data analytics to identify anomalies and defects before failures occur that require lengthy downtimes, to maintain/repair machines in a timely manner when needed rather than at multiple times each year, and to avoid replacing parts that are still working),
• prediction of the remaining useful life of complex systems (modelling the lifespan of machine components to avoid serious unplanned outages, unnecessary preventative maintenance, and other failures in the overall manufacturing system);
• product quality monitoring, amongst others.

Dr Ali Intizar, Assistant Professor at DCU and the Insight Centre, said “In the manufacturing domain, limited or partial data availability often leads to incomplete knowledge bases. Our work addresses this challenge by predicting links within knowledge graphs, thereby completing industrial knowledge graphs and enabling the development of smart applications for manufacturing production lines”.

Prof. John Breslin of the Insight Centre at the University of Galway, said “This work highlights the need for Knowledge Graphs, which requires significant effort in the development of standardised semantic models, methodologies, and tools for not just sports production lines, but for multiple industrial sectors and applications.”

The Insight SFI Research Centre for Data Analytics at the University of Galway is a founding member of the Knowledge Graph Alliance, a non-profit organisation in collaboration with Bosch and 20 other organisations (universities, companies, industry consortia), launched in November 2023. The board consists of a mixture of industry members [from TotalEnergies, Airbus, Michelin] and academics.

This research paper is available at: https://doi.org/10.1109/ACCESS.2024.3419911

 More information on the Knowledge Graph Alliance is at: https://www.kg-alliance.org/

 The Insight SFI Research Centre for Data Analytics is now hiring for a postdoctoral researcher on Knowledge Graphs to contribute to this research space: https://universityvacancies.com/university-galway/postdoctoral-researchers-x2-insight-and-vistamilk-1-fte-data-science-institute