× close
Detect heat loss hotspots from multifamily buildings using deep learning with bounding boxes.Credit: University of Waterloo
Researchers at the University of Waterloo have developed a new method that could lead to significant energy savings in buildings. The research team identified 28 major heat loss areas within multifamily buildings. The most severe heat loss areas were around wall intersections and windows. If 70% of the discovered area is remediated, a potential energy savings of 25% is expected.
Their research paper is Energy conversion and management.
Building enclosures rely on heat and moisture control to avoid significant energy losses due to airflow leaks. This makes the building less comfortable and more expensive to maintain. This problem is likely to be exacerbated by climate change due to erratic temperature fluctuations. Manual inspections are time-consuming and infrequent due to a lack of trained personnel, making energy efficiency a widespread issue for buildings.
Waterloo researchers, a leader in sustainability research and education and a promoter of environmental innovation, solutions and talent, have created an autonomous, real-time platform to make buildings more energy efficient. The platform combines artificial intelligence, infrared technology, and mathematical models that quantify heat flow to better identify areas of heat loss within buildings.
Researchers used a new method to conduct an advanced study in a multifamily building in the extreme climate of the Canadian Prairies. There, elderly residents reported discomfort and higher electricity bills due to increased demand for heating in their units. Using AI tools, the team trained the program to inspect thermal images in real time, and he achieved 81% accuracy in detecting areas of heat loss within building envelopes.
“The nearly 10% increase in accuracy with this AI-based model is effective because it improves occupant comfort and lowers utility bills,” said Waterloo, director of the Architectural Engineering Program. said Dr. Mohammad Alazi, Director of the Symbiotic Research Institute. It is an interdisciplinary group of universities specializing in the development of innovative building systems and the construction of greener buildings.
New AI tools help remove the element of human error when examining results, increasing the speed of data analysis by 12 times compared to traditional building inspection methods.
In the future, we will expand this initiative by using camera-equipped drones to inspect high-rise buildings.
“Our methodology can be used to analyze buildings and is expected to lead to millions of dollars in energy savings in a much faster manner than previously possible,” Arazi said.
For more information:
Ali Waqas et al., Machine Learning-Assisted Thermography for Autonomous Heat Loss Detection in Buildings, Energy conversion and management (2024). DOI: 10.1016/j.enconman.2024.118243
