Machining Intelligence Network: Data Mining and Semantic Search in Manufacturing Industry

Ines Färber, Sergej Fries, Götz Marczinski, Thomas Seidl, Nils-Per Steinmann

Abstract: The “Machining Intelligence Network” (MachInNet) project tackles the challenges to “unearthing” manufacturing knowledge from NC codes (numerical control codes), tool layouts and other manufacturing documents, and to making it accessible for daily use, e.g., for feature-based NC planning. A new approach using data mining algorithms and semantic search technologies makes it possible to reverse engineer data from different sources and make it available for explicit use with the help of a semantic, Internet-based knowledge network. The business rationale of MachInNet is to help SMEs (small and medium-sized enterprises) to manage a large variety of technologies, to avoid redundant engineering efforts, and to accelerate industrial engineering of mechanical parts.