DEVELOPMENT OF AI-BASED ENGINEERING BIG DATA INTEGRATED ANALYSIS SYSTEM FOR DECISION-MAKING SUPPORT IN THE ENGINEERING-PROCUREMENT- CONSTRUCTION (EPC) INDUSTRY

Abstract

Plant Engineering-Procurement-Construction (EPC) industry is one of the complex industries going through various stages from bidding to engineering, construction and operation and maintenance (O&M). A systematic management system is needed to address these complexities. However, many EPC companies in Korea are having difficulty managing their projects due to the lack of data-based systematic decision-making, and are suffering heavy losses in overseas projects. The AI-based engineering big data integrated analysis system proposed by this study aims to minimize project losses and eventually to enhance the technical skills and competitiveness of the Korean plant industry through decision-making support, combining big data and AI in the entire EPC project life cycle. In this study, knowledge base was established to utilize various data generated during the entire EPC project life cycle in AI-based engineering big data integrated analysis systems. And a machine learning integrated platform specialized in the engineering industry was developed to support feature engineering, model learning and model operation processes. Using various algorithms from the machine learning integration platform and the knowledge base, five main decision-making applications were developed: analyzing bidding documents, predicting design costs, analyzing design errors, analyzing change order, and plant equipment prediction maintenance. Based on the predicted information, the system could help EPC project managers identify and manage risks at each stage of the project in advance to make decisions that minimize project loss. Furthermore, the information predicted at each stage may be circulated or used as feedback for decision making at other stages.