Digital shop floor management enhanced by natural language processing

Abstract This paper aims to develop concepts how digital shop floor management (dSFM) can be further enhanced by natural language processing (NLP) to bring a higher value to the shop floor team and decision makers. Based on the literature review on these two fields several valuable application of NLP in dSFM are theorized: recommender engines to improve knowledge management, text clustering to identify frequent problems, voice assistants to ease the interaction with the data base, chat log extraction to fill the database with unstructured written text from chats and spellcheck as well as auto fill to improve data quality. To show the feasibility for NLP in dSFM in industry, a case study for the document clustering is presented: A digital ticket system for shop floor issues used for two years and containing 2,735 entries is analysed with the “Graph”-feature from Elasticsearch to find the most frequent terms and intersections in the described problems. The approach is accurate, quick and detailed and will be established in the company and performed monthly.