The research focuses on overcoming the limitations of current BEMS by integrating and extending existing ontologies, specifically BRICK and KPIOnto, to form the NESTKPI ontology. The extended ontology bridges the gap between BRICK and KPIOnto to capture the complexities of KPI representations within the BEMS context. Another aspect of this work is the development of the Function Layer, a collection of intermediary scripts designed for handling data processing from time series databases and integrating it into the knowledge graph. This layer is important for ensuring accurate and efficient KPI calculations.

The validation of the proposed approach involves careful testing through real-world use cases and expert feedback, confirming its effectiveness in enhancing KPI management.

The thesis demonstrates the potential of knowledge graphs in improving the efficiency and scalability of diagnostic analyses of BEMS, offering a contribution to the field of building energy management systems and setting a foundation for future advancements in this domain and KPI integration.