DATAWiSE: A Scalable Big Data Reference Architecture for Smart Building

The emergence of Digital Twin technology has dramatically changed operational and managerial practices in intelligent buildings by enabling real-time monitoring, predictive maintenance, and performance improvement. However, the management of large and heterogeneous data produced by Internet of Things (IoT) sensors, Building Information Models (BIM), and environmental systems poses a challenge in terms of interoperability, scalability, and real-time processing capability. This paper presents the DATAWiSE architecture, a modular and scalable big data framework that is designed to integrate disparate building data sources while ensuring both secure and efficient data management. A detailed account of the process followed, from user requirements collection to the design of the proposed reference architecture. The proposed multi-layered architecture combines distributed storage technologies, AI-based analytics, and standardized communication protocols to provide strong data governance while enabling informed decision-making. Explainable AI approaches also assist in increasing system transparency and stakeholder trust. Future development areas include further enhancements in federated learning, edge computing, and harmonization with international data-sharing protocols to increase the architecture's applicability within larger smart city infrastructures.