U.S.-Ireland R&D Partnership: Intelligent Data Harvesting for Multi-Scale Building Stock Classification and Energy Performance Prediction

Project: Research project

Project Details


This joint trilateral project is in response to NSF Dear Colleague Letter: United States- Ireland-Northern Ireland R&D Partnership (NSF 20-064) in the areas of energy and sustainability. Residential buildings account for 14%-27% of greenhouse gas (GHG) emissions in the three jurisdictions and cause significant negative impact on the environment. Supported by NSF, the Science Foundation Ireland in the Republic of Ireland (RoI), and the Department for the Economy in Northern Ireland (NI), this joint research aims to reduce residential building energy consumption and related GHG emissions and environmental impacts across the three jurisdictions. The research will create decision support tools to inform policy makers, planners, and other stakeholders about the most beneficial residential retrofitting solutions at multiple scales (local to national). The methodology employed will lie at the confluence of various expertise, including green engineering of the NI team, building energy modeling and machine learning of the U.S. team, and information theory of the RoI team. The aim is to transform diverse public datasets in the three jurisdictions into actionable information. Empowered by this information, the anticipation is that better decisions can guide modern societies towards transformative green solutions for the built environment that leverage sustainable engineering systems and enable the creation of energy-efficient, healthy, and comfortable buildings for a nation's citizens. The approach is cognizant of society's need to provide ecological protection while maintaining favorable economic conditions.This joint research seeks to provide the foundational science needed to design, optimize, and deploy green engineering approaches that reduce residential building energy consumption and related GHG emissions. The interdisciplinary research targets to yield three results: 1) A methodology for data ingestion and an ontology and associated server that provides both a means of accessing and subsequently homogenizing data for both the data enrichment and the modeling processes. The intent is to enable previously unused data sources to be utilized as a whole to significantly improve the accuracy of modeling processes; 2) An advanced automated building energy model generation method powered by physics-informed machine learning, which can improve the efficiency of model generation, significantly reduce computing demand for large scale building energy prediction and protect building users' privacy. Algorithms will also be created to enable robust prediction with incomplete datasets; 3) A new complementary solution for predicting the GHG emissions reduction potential for stakeholders will be created to analyze near/zero GHG buildings in terms of energy performance. It is anticipated that these results will be beneficial both in terms of making buildings greener by reducing GHG emissions and energy consumption as well as decreasing operational costs. The plan is to seek the U.S. Department of Energy's Pacific Northwest National Laboratory to adopt the research results in their national building energy policy analysis for 139 million homes. The Northern Ireland Housing Executive will utilize this work to help predict decarbonization pathways for their housing stock of nearly 86,000 homes (10% of the housing stock in NI). The research will also assist the Sustainable Energy Authority of Ireland for its retrofit plan of 500,000 homes in the Republic of Ireland.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
Effective start/end date1/1/228/31/25


  • National Science Foundation: $388,623.00


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.