A web-based approach to BMS, BIM and IoT integration

a case study

Abstract

Buildings are complex cyber-physical systems that rely on a combination of heterogeneous systems to provide smooth operation, energy efficiency, occupant comfort, well-being and safety. Building Management Systems (BMS) are central to these operations and generate a huge amount of data. Traditionally operated in a local server in a building, state-of-art BMS solutions are now moving towards the cloud. Internet of Things (IoT) meters and sensors are also increasingly used in buildings and hold tremendous potential for smart building monitoring and control. Another valuable data source is a properly developed and managed Building Information Model (BIM). BIM itself is moving towards Level 3, which is web-based and data-driven, as opposed to the file-based BIM of today. Improving building performance (e.g., energy, comfort, operational cost) relies on data from all the above systems. However, these data usually remain siloed within their own environment and do not provide an opportunity to perform evaluations across multiple systems. Moreover, valuable information about geometry, spatial location, and metadata about the building objects that the BIM models already contain remains unusable for building performance monitoring and reporting tools. Integrating these heterogeneous data sources will provide ample opportunities to improve building performance. Even though some commercial tools enable the integration of sensor data with BIM models, such tools remain largely proprietary and are not compatible with other applications. This study presents a methodology to integrate multiple information sources at their system level in a distributed manner. The Industry Foundation Class (IFC) model of the building is used as the primary source of information for creating a semantic building graph. Since the semantics of BMS sensors was not originally available in the IFC model, the Brick ontology is used to semantically describe BMS sensors in the graph. Sensor data related to spaces in the BIM model is visualized by selecting a space from the 3D model via the web application. Each data stream remains in its optimum environment and the connections are made via an Application Programming Interface (API).