TU/e Eindhoven University of Technology
OPERATIONS PLANNING ACCOUNTING & CONTROL GROUP
DEPARTMENT OF INDUSTRIAL ENGINEERING & INNOVATION SCIENCES
Abstract
Over 70% of data in organisations is unstructured text and engineers spend 14-30% of their time searching for information. State of the art information extraction tools present an opportunity to significantly improve engineers' productivity. However, engineers are reluctant to rely on the outputs of neural language models alone. Checks need to be made to ensure the rules of physics and engineering common sense are observed. This talk describes work by the authors using NLP and ontologies and how these two fields are being integrated to enable reasoning for widely used engineering processes. The talk focuses on the use of industrial NLP, ontologies and knowledge graphs on equipment maintenance work order records, a key asset in improving product life cycle and asset management performance.