Iowa State researcher develops AI technology to optimize solar installations

CATEGORIES: News, Notable, Research
A large array of solar panels

As solar power becomes more accessible, affordable and popular among U.S. homeowners, Forrest Sheng Bao, assistant professor in the Department of Computer Science is developing artificial intelligence (AI) solutions that will optimize residential solar installations—from the initial price quote to the final tile placement.

RisingSun Solar, a growing solar company headquartered in Kansas City, Missouri, will use the AI technology to configure system-design options without an installer stepping on a roof. Price quotes will be generated in seconds, without the hassle of in-person meetings.

“Iowa State’s AI tools are enhancing our effort to develop the first digital online marketplace where homeowners can purchase custom solar solutions that are tailored to their home and specific energy needs,” said Keith Murphy (’11 marketing, business management), chief executive officer at RisingSun Solar. “With just a couple of clicks, we make it very easy to show customers what it looks like to power their home with the sun.”

AI meets solar: How does it work?
Bao’s AI technology, an application programming interface (API), will run on RisingSun Solar’s front-end sales website. After a customer enters their address and most recent electric bill, the API accesses satellite images of their home via Google Earth. A machine-learning algorithm calculates the size of the home and shape of the roof, based on the image data. The API also considers the roof orientation as well as the longitude and latitude of the city where the home is located.

In seconds, the API reveals if solar energy is feasible for that particular home and the number of solar tiles needed. Future iterations of the API will display renderings of solar-design options, from which customers can pick and choose.

Because of the labor and time savings, the API will drastically cut the non-hardware costs of going solar.

“This leads to more people going solar, which is good for the environment and good for our company,” Murphy said. “These AI-based systems are completely turning the current solar sales and marketing process on its side.”

AI: A bright spot in solar energy
Bao, a skilled AI researcher who specializes in AI in electronics design automation, agrees that AI is quickly evolving and rapidly transforming business practices across a wide variety of sectors.

“Artificial Intelligence keeps getting hotter and is beyond what I could have imagined 10 years ago,” Bao said. “Most AI projects that we’re developing in the computer science department, including the API for Rising Sun, use technology that didn’t exist a decade ago.”

RisingSun Solar’s Chief Technology Officer, Eric Andren (’12 graphic design, ’14 MFA), notes that Bao’s AI technology will be integrated at multiple junctures in the company’s “A-to-Z process of going solar,” and will automate how it markets products to its solar-seeking customers.

“The AI that we’re using for price quotes cuts a day or two from the process,” Andren said. “But there’s an entire ecosystem of AI products that Bao will be developing, which will reduce a two-month solar installation process to a few days or even hours.”

Iowa State University: A hub of AI innovation
Bao’s AI technology is one example in Iowa State’s deep bench of artificial intelligence capabilities.

“Computer science and data science faculty have a wide range of AI expertise and talents that have made significant impacts in applied AI—like the solar tile project. Many are also accomplished AI researchers and scholars who run their own labs,” said Hridesh Rajan, Kingland Professor of Data Analytics and computer science department chair.

In June, the Department of Computer Science launched the first AI graduate program in Iowa, further solidifying Iowa State as hub of AI innovation, research and scholarship. The two-year master’s of science degree prepares students for high-demand AI careers in machine learning, deep learning, signal image processing, natural language processing, deformable objects, data-driven security, automated processes using robots and others.

“It’s an exciting time for the Department of Computer Science as we build on strong foundations in AI, computer science and data science,” Rajan said. “Not only are we providing opportunities for students to participate in cutting-edge research, we’re also developing real-world solutions for companies like RisingSun Solar, which is in line with Iowa State’s central mission as a land-grant university.”