Notre Dame researchers developing dashboard to help low-income households save energy
Subscriber Benefit
As a subscriber you can listen to articles at work, in the car, or while you work out. Subscribe Now
A grant from the National Science Foundation has enabled University of Notre Dame researchers to employ data and analytics to find energy cost-saving solutions and improve living conditions for low-income Americans living in older, less efficient homes.
The BUILT2AFFORD initiative is a collaboration among professors in the school of architecture, the college of engineering, the Lucy Family Institute for Data & Society and the Fitzgerald Institute for Real Estate.
The goal is to develop, test and validate a tool that uses Google Street View and machine learning to identify houses suitable for low-cost retrofits.
“Think of a building as a human body, the human body has some very weak points. For example, the neck and ankle. If you just put a scarf around your neck and wear thick socks, without wearing a heavy jacket, you will probably increase your comfort level,” Associate Professor of Architecture Ming Hu said. “The building can function in a similar way, so insulating the entire building does not necessarily lead to the best optimized performance. The best thing we can do is to insulate the critical parts, cutting off the thermal leak at specific places and also reduce unnecessary cost.”
In lower-income neighborhoods especially, Hu said older homes tend to suffer from inadequate insulation, leaky ducts, cracks, drafty doors and windows and outdated HVAC systems leading to higher energy costs, compromised air quality and uncomfortable living conditions.
The BUILT2AFFORD project uses computer vision models and Google Street View images to identify physical characteristics of buildings that affect energy use and indoor thermal comfort. Information from open-source databases are also included to form the foundation of the prediction tool.
“It started with the idea of what we could do to get ahead of this problem of climate change for an area that’s in some ways buffered from it,” Matthew Sisk, associate professor of the practice of data science, said. “We’re not going to get massive tidal surges or hurricanes in the Midwest, but we are going to have changing temperatures and that has huge impacts on people.”
Sisk, who has lived in South Bend for over a decade, also brings his experience living in a home that is almost 140 years old to this project. Sisk leverages his lived experience, his data science background and the relationships within the neighborhood to lead community-facing research projects at the Lucy Institute.
“The housing stock in northern Indiana was built for a climate that we had 50 to 100 years ago,” Sisk said. “The climate has changed, so understanding some low-cost ways that we can make these houses more efficient and make them a little bit more useful for people to live in is the overarching goal.”
The group has also teamed up with several community partners, including the city of South Bend, South Bend Heritage Foundation, Oldtown Capital Partners and the Near Northwest Neighborhood (NNN) to conduct tests on a variety of housing types in an economically diverse area.
“These are some of the neediest of houses. The houses themselves are around 120 years old, so there’s a lot of drafty homes, there’s a lot of energy and efficiency concerns, so we’re really excited to have this partnership,” Near Northwest Neighborhood Executive Director Kathy Schuth said. “This neighborhood gives them the grounding that they need for a really good pilot program. We can do the work of connecting them to neighbors in a holistic way.”
The team has conducted several field audits where they create a 3D scan of the home, take thermal images, measure for moisture content and document building materials and insulation. Sensors to measure temperature, humidity and air quality are then placed around the house for one week. The results help the team to hyper-locate problem spots and create an exposure model to predict risk across housing types.
“We’ve been surprised that some homes that have had full rehabs actually have worse air quality, and I find that very interesting,” Schuth said. “I’ve really appreciated the targeted approach they’re taking—trying to see if there’s one room in the house that is the culprit. You could add insulation to one corner of the house instead of the whole house and really make a strong difference.”

Notre Dame has worked with the NNN on several projects, including GIS mapping and reducing lead exposure among children in the neighborhood.
“We connected with Oldtown Capital through the city. They gave us access to Marmain Apartments for some tests,” Hu said. “We are looking into low-cost affordable housing, and Marmain has 46% of Section 8 units. Those are units specifically reserved for low income households.”
Hu said that while old buildings often lack good insulation, the thermal retention in most of the aged buildings is brilliant, especially as new houses often do not perform better. So instead of spending a whole lot of money, low-cost solutions like air sealing, new windows, upgraded ventilation systems and exterior shading could make a huge difference.
Just from sealing windows, Hu said she observed a 30% reduction in air leaks during the simulations. For those who can afford to install a new window, Hu said there are two major characteristics you need to look out for.
“The first thing is the Solar Heat Gain Coefficient, a ratio that measures how much solar heat passes through, and the second is the Visible Light Transmittance, a measure that shows how much daylight can get in,” Hu said. “So it’s always a trade off in the winter, especially here in our region. You want less solar radiation leakage, but you want a lot of light.”
For Schuth, this project has helped families in the Near Northwest Neighborhood access information that would normally be out of reach. And unlike an insurance salesperson offering energy efficiency solutions, Schuth said it felt good to know there was no catch at the end of the simulation.
“This kind of collaboration is a real, beautiful offering. They’ve allowed us to bring the community to it and they’re bringing so much expertise, data, and science in a field that we don’t usually have access to,” Schuth said. “We’ve really found it to be a very solid collaboration that’s really authentic and good. It’s made a very daunting thing become very accessible.”
To validate their research, Hu and her team, in partnership with the Fitzgerald Institute for Real Estate are already applying for a Phase 2 NSF grant. The grant would pay for at least 80% of the cost of the retrofits, with local programs potentially covering the rest. If they receive the award, the solutions will be implemented next January and results would be available by December 2026.
As part of its Greener Homes, a pilot program for low-income, elderly homeowners who are part of Habitat for Humanity of St. Joseph County’s Aging in Place Program, the city of South Bend currently pays for some energy upgrades in partnership with enFocus.
“We are hoping the dashboard we develop can be included in this program for greener homes,” Hu said. “Another initiative by the city requires local citizens to do energy audits as a part of the package to apply for a grant. With our open access dashboard, anyone can generate the energy assessment report to be included in the application package.”