Analysis and assessment of Maintenance Logs: designing an AI-enhanced integrated system to improve data-driven decision-making

This thesis was conducted at Dumarey Flowmotion Technologies, an automotive company specializing in the design and production of injectors for gasoline engines and exhaust gas treatment systems for Diesel engines. The project is exploratory in nature and analyzes data from the DEKA 4/5 plant, with a particular focus on machine breakdowns that impact Overall Equipment Effectiveness (OEE), maintenance costs, and workplace health and safety. The work is structured into three main phases: (1) Mapping the AS-IS state of maintenance data, along with an assessment of the quality and relevance of the problem hypothesized by the company. (2) Developing a data processing workflow focused on extracting information from textual data related to machine breakdowns, highlighting the strengths and weaknesses of the current system. (3) Designing a new model system to which the company refers for transitioning to a more advanced system, supported by artificial intelligence (AI), particularly Natural Language Processing (NLP) algorithms. The proposed system is integrated with FMECA (Failure Mode and Criticality Analysis), Reaction Plan, WMS (Warehouse Management System), Shopfloor, Audit Controls System, and Production Sheets.