Semi-automated knowledge graph construction for cross-platform interoperability in building IoT systems
Junsong Du, Max Berktold, Jinzhi Jiang, Yunheng Tian, Rita Streblow, Dirk Müller · Results in Engineering · 2026
Automating the deployment of control and monitoring applications is essential for achieving energy-efficient and intelligent building operations. Internet of Things (IoT) platforms provide a powerful foundation for this by enabling seamless data access, but their widespread adoption is hindered by a lack of cross-platform interoperability. While Semantic Web Technologies offer a theoretical solution, their practical application is limited by the significant manual effort required to formalize engineering information into Knowledge Graphs (KGs). This paper addresses this gap by presenting a novel method that semi-automates the KG construction from heterogeneous IoT platforms and leverages these KGs for the automatic deployment of controller applications. A key feature of this approach is the unified modeling of both building-specific information alongside platform-specific API interaction mechanisms within a single graph. We validate its effectiveness in a comprehensive case study involving heterogeneous test environments. The results confirm that the method can create reusable pipelines that automates the generation of complex KGs, thereby enabling the fully automated deployment of controller applications across heterogeneous building IoT systems. These findings provide a practical path toward scalable, cross-platform interoperability in the building sector.