Grant: Discovery Grants Program - Individual

Buildings are a critical component of Canada's infrastructure and economy and play a major role in energy use (40% of total), GHG emissions (33% of total), and occupant health and productivity. The correction of operational inefficiencies (e.g. poor controls logic, faulty equipment, etc.) can provide an estimated 20-30% energy savings in each building with long-term economic and environmental benefits. Facility Management (FM) practices are beginning to leverage Smart Buildings technologies to leverage sensor networks and integrate multiple systems to address resilience, emergency preparedness, occupant health and productivity, and reduce energy and CO2 impacts. New tools, insights, and strategies are necessary to support this transition and enable Canadian buildings to achieve new levels of operational efficiency. The proposed research program leverages the PI's previous research in energy analytics, Facilities Management (FM) enabled Building Information Modeling (FM-BIM), and integrating machine learning into FM system. Through long-term collaboration with the FM team at Ryerson University, two buildings under construction (opening 2018) have been outfitted with sensor networks and will stream to the Virtual Campus Model previously created. A third building to be completed in 2019 will host the new Smart Building Analytics Lab opening in 2019) and provide a living lab to test and demonstrate new controls and operations strategies. The proposed research will start with the development of a data model expanding existing IoT and BIM ontologies to support Smart Continuous Commissioning. Data obtained from the building sensor networks, and institutional facility management systems will be mapped using this structure into a Building Information Model (BIM) to test the robustness of this model and permit spatio-temporal visualization. Physical and virtual testbeds will be used to prototype, develop, and test data retrieval, mining, analysis, and management at the IoT devices using both supervised and unsupervised machine learning approaches. The PI's new Smart Analytics Laboratory along with a campus FM-BIM (to both provide facility data to the IoT devices and visualize the near-real-time performance) and Matlab building emulator will serve as the physical and virtual testbeds for experimentation. Over the five-year duration of the grant, it is expected that this research will lead to novel data integration and management approaches based on the new data model, new fault detection, trending, and analysis techniques, and a Smart Continuous Commissioning (SCCx) system prototype. This research will reduce the barriers to continuous commissioning, facilitating the maintenance of peak operational performance within buildings with the potential to significantly decrease energy consumption and associated greenhouse gases.