Computation of semantic similarity between products by analysing data collected by PLM and MES systems

The extensive adoption of digital technologies by manufacturing companies, imposed by the global competition and the advent of industry 4.0, has led to the generation of huge amounts of data. If well exploited to create useful knowledge, such data can provide the companies with superior and sustainable competitive advantage. For this reason, there is an increasing interest in manufacturing companies to develop advanced knowledge management systems. Among different companies, those operating in one-of-a-kind production find it crucial to effectively manage their knowledge because they are under the pressure of delivering high quality products in a short time and at low cost. To this aim, several researches have been conducted to effectively manage the knowledge and most of them highlighted the importance of integrating information coming from different sources. This thesis addresses the problem of identifying similar products by proposing a knowledge management framework that exploits the data integration between Product Lifecycle Management (PLM) and Manufacturing Execution System (MES). It focuses on the application of semantic measurements that use ontologies to calculate product similarities. The effectiveness of semantic similarity method through the adoption of information content approach is demonstrated by applying it on real data belonging to a company that produces prototypes components in the automotive sector and its products are highly customized and should match with individual customer needs. The aim is to capture the implicit knowledge embedded inside employees minds and make it transferable and reusable by other employees to enhance the way of defining production cycles for new products and to make them more accurate.