UNIVERSITY PARK, Pa. — Wangda Zuo, Penn State professor of architectural engineering, is part of a team of researchers recently recognized by the journal Energy and Buildings for work that “withstands the test of time.”
The paper for which Zuo and his team were recognized was published in 2024, titled “Urban Building Energy Performance Prediction and Retrofit Analysis Using Data-Driven Machine Learning Approach.” The work details a scalable artificial intelligence (AI)-based framework that uses the physical parameters, such as space and time of day, of urban residential buildings to predict their energy performance. The framework can adjust to assess how various retrofits, such as installing energy-efficient windows, may impact the building’s overall energy use.
Energy and Buildings is an international journal “devoted to investigations of energy use and efficiency in buildings,” according to their website. This is the third time the journal has awarded “best paper” awards, with this round recognizing five research papers published in the journal in 2023 or 2024. Previous awards were made in 2023, for papers published 2018-22, and 2019, when the editors recognized papers published 1998-2017.
“We believe the ‘best papers’ for Energy and Buildings should withstand the test of time, which can be partially reflected by their frequency of citations, but also the novelty and impacts they demonstrate,” the journal’s editors-in-chief, Jian-lei Niu and Mat Santamouris, wrote in an announcement.
According to Zuo, the goal of the researchers' project is to ultimately inform policymakers and urban planners developing strategic sustainable energy plans to reduce energy consumption and emissions from the built environment.
“Our team has worked closely with stakeholders to understand the challenges they face in improving building energy efficiency,” Zuo said. “We developed building energy models, created methods to process large datasets describing building stocks and explored machine learning approaches for predicting building energy use. This paper brings these pieces together. By combining physics-based building models with artificial intelligence, we developed a ‘physics-informed AI’ approach that can predict building energy performance more accurately and efficiently, helping support better planning and operation of energy-efficient buildings and communities.”
This paper was part of a broader, ongoing research project funded by a four-year, $1.2 million U.S.-Ireland collaborative project jointly supported by the U.S. National Science Foundation, Science Foundation of Ireland and the Department for the Economy in Northern Ireland.
“This recognition highlights the growing role of physical AI in transforming how we design and retrofit buildings' energy systems,” Zuo said, referring to the idea of artificial intelligence that can understand and adapt to spatiotemporal relationships in the real world. “By integrating physical principles and physics-based models with machine learning, we can deliver more sustainable and affordable housing solutions. It is especially meaningful to see this U.S.-Ireland collaborative research recognized, as it contributes to addressing real-world energy challenges in the built environment.”
Through the research process, Zuo and his collaborators have worked closely with government officials, industry professionals and national labs to gather insights and feedback on their work. The team hosted a U.S.-Ireland workshop on multi-scale building energy modeling in Dublin in May 2024, with more than 20 stakeholders from three jurisdictions attending. The workshop focused on understanding barriers to both research and real-world deployment of energy-efficient technologies, and on identifying future research directions that could accelerate the transition toward more sustainable and energy-efficient buildings.
The paper was co-authored by Usman Ali, Sobia Bano, Divyanshu Sood, Cathal Hoare, James O’Donnell, all with University College Dublin; Neil Hewitt, Ulster University; and Mohammad Haris Shamsi, Flemish Institute for Technological Research.