ABSTRACT

More than 125 million residential and commercial buildings in the United States account for 40% of primary energy use and 73% of electricity use on the nation’s grid (80% during peak demand); the total energy cost of operating U.S. buildings is $370 billion per year. Many home or building owners have been faced with an urgent need to (1) replace failed heating, cooling, and other equipment, (2) improve their building energy efficiency or add off-grid electricity amidst confusion over actionable options, or (3) clarify cost-effective priorities for optimal retrofit or financial investment when the information available is often inadequate. Scientists at ORNL are leveraging high-performance computing resources to simplify building-specific decisions and to provide major industry partners with data necessary to increase deployment of energy-efficient and other building technologies to help meet national goals.

After five years of effort, the team I led successfully created a model of every U.S. building in 2020. The 122 million models were created by mining imagery (satellite, aerial, street-view);

employing light detection and ranging (LIDAR) technology, which can scan large volumes and rapidly collect precise three-dimensional data; using multiple listing services, and developing databases for buildings to consolidate digital information on the unique footprint, height, age of construction, and other properties necessary to create a building energy model in an open-source tool. 

Each U.S. building was then simulated with typical weather for that area to estimate baseline energy use and validate the results against utility data. This work achieved several breakthroughs in the field of “urban scale energy modeling” via the introduction of novel, scalable data, as well as the theoretical and practical advancement made possible by artificial intelligence, computer vision, and big data processing. The digital twin data for 125.7 million U.S. buildings and 122 million models have been made open-source (bit.ly/ModelAmerica1), and software has been developed to help inform best practices for urban-scale energy modeling (bit.ly/AutoBEM).

In the past five years, I have been working with key partners in several multibillion-dollar industries to enhance the data and models for practical uses in actionable business decisions. My collaborators and I work with (1) architecture, engineering and construction firms, to inform energy-award decisions during early design, (2) local power companies, to make operational decisions on their electrical distribution network, (3) energy service companies), for preliminary audits of performance contracts, (4) commercial real estate lenders, for future operational costs of commercial buildings, (5) hardware and home improvement stores, for third-party assessment of savings for equipment or building retrofits, (6) software companies, to improve tools and estimate the energy impacts of over 40,000 cities, and (7) federal agencies, for prioritizing investment of taxpayer resources for making building improvements. Yet, even more exciting multitrillion-dollar opportunities exist for resilience, insurance, and automated financing of building investments.

In my talk, I will present some of the national goals within the U.S. Department of Energy (DOE), scientific breakthroughs and the roadmap for how to create a digital twin of every building in a nation, ways this information is being used currently, and some thoughts regarding how it could help scale up existing industries or create new ones.

BIOGRAPHICAL SKETCH

Dr. Joshua New is a distinguished R&D staff member at Oak Ridge National Laboratory. He is also a joint faculty member in the Electrical Engineering and Computer Science Department and Bredesen Center at the University of Tennessee at Knoxville, where he received the Outstanding Mentor Award in 2024 and his Ph.D. in computer science in 2009. 

In 2024 he received the international R&D 100 “Researcher of the Year” award. In 2016 he won an R&D 100 award (Oscar of Invention) titled “Roof Savings Calculator Suite.”

Dr. New has more than 200 peer-reviewed publications. He managed research portfolios for the Building Technologies Office that totaled more than $25 million while he was on loan to DOE. Over the past 15 years, he led more than 225 projects totaling $350 million.

His expertise involves building energy modeling, supercomputing, AI, and big data mining. This knowledge and skill set have been used to generate, simulate, and analyze every U.S. building’s energy data. The data has been made publicly available for scalably quantifying energy use, energy demand, emissions, and cost reductions from individual buildings to the entire nation.