Development of ontology based knowledge construction tool with large language model

Life Cycle Engineering (LCE) is a critical paradigm for manufacturers striving to implement environmentally sustainable and resource-efficient practices. Reliability-Centered Maintenance (RCM) contributes significantly to key performance indicators such as equipment uptime and environmental impact. However, its implementation heavily depends on expert knowledge derived from Failure Mode, Effects, and Criticality Analysis (FMECA), which requires substantial effort and rigorous quality control. To address these challenges, we have developed three key methodologies, the construction and automatic evaluation technologies of asset knowledge graphs using large language models (LLMs), graph database technologies, and human-in-the-loop evaluation technologies. These technologies support the ingestion of custom ontologies defined in RDF format, enabling flexible and scalable knowledge graph generation across diverse domains. Furthermore, the generated graph data can be transformed into human-readable tabular formats to enhance interpretability and reviewability.The proposed framework streamlines ontology-driven knowledge construction and evaluation, improves the efficiency of knowledge management processes, and enhances overall productivity in maintenance engineering.