부산대학교 도서관

As the Internet-of-Things (IoT) grows and data analytics mature, their application to new buildings is proved a significant opportunity for improved building performance at reduced energy costs. In existing buildings, however, the cost to replace existing Building Management Systems (BMS) with IoT-compatible devices poses a barrier to adoption. Additionally, the limited data storage on these existing systems further precludes data analytics for optimization. This research responds to this need by presenting a novel approach to pre-process and stream the BMS data to a cloud-based database on a private network. Afterwards, in the centralized data warehouse, larger scale and more complex analytics can be performed. This paper presents the development of both the new database architecture and supporting infrastructure to support the streaming of BMS, as well as the pre-processing to optimize big data analytics and visualization. A proof-of-concept visualization for a 14,000m 2 student learning center is presented to demonstrate the application of this architecture.