AI in Construction Institute
Faculty members team with more than 80 organizations to foster new, high-impact research in the construction industry.
Given the in-roads that Grainger Engineering computer vision and artificial intelligence experts were having in the construction industry, they knew it was time to have even greater impact. So in 2021 they launched the development of the AI in Construction Institute. It will develop the ideas, technologies, and workforce needed for the design, construction, and maintenance of buildings and infrastructure.
The institute has received funding from the National Science Foundation and the University of Illinois System’s Discovery Partners Institute, which supports training, applied research, and development of the technology businesses in Illinois to plan the institute. More than 80 organizations are also involved. They include construction companies and architecture and engineering firms like Clayco and Skidmore, Owings & Merrill, massive real estate owners like Amazon, technology companies like Autodesk, Oracle, and Facebook, and venture capital investors. Faculty from Carnegie Mellon University, a key academic partner, are also participating.
Leading the institute at Grainger Engineering are: Professors Mani Golparvar (Civil & Environmental Engineering) and Derek Hoiem (Computer Science), as well as Professors Randy Deutsch (Architecture), Nora El-Gohary (CEE), Khaled El-Rayes (CEE), David Forsyth (Fulton Watson Copp Chair in Computer Science), and Julia Hockenmaier (CS), and E.J. Ignacio (CEE).
“We’ve tapped into this mass audience in the construction industry that really wants to be involved in basic research. They want to be engaged in bleeding-edge research and education,” said Golparvar. He expects research in areas like computer vision, natural language processing, and machine learning — all of which require huge quantities of annotated data to develop, test, and train AI algorithms.
This data has been collected in some fields like advertising and autonomous vehicles. But construction lags. Accordingly, developing shared data resources for organizations that are members of the institute will be an early priority.
“Researchers tend to gravitate towards problems where there’s readily available data because they don’t want to deal with a data collection issue. That leaves large swaths of opportunity unexplored,” according to Hoiem. “In order to make headway on these problems, you need to get access to data that only construction and construction tech companies have.”
That’s why the Construction in AI Institute’s long list of partners is so important. They have the data, and the institute and The Grainger College of Engineering can provide a space where companies can share data sets that they can use individually to improve their proprietary technology and to conduct collaborative research.
“Data is the most important thing to figure out, especially if you’re doing research in AI,” Oracle’s Burcin Kaplanoglu. Oracle is a major partner in the Construction in AI Institute, and Kaplanoglu also chairs a committee that advises the institute. It includes leaders in construction, architecture, and engineering as well as tech providers, startups, the investment community, and owners and operators of the built environment.
“The key is mutual — having a neutral place to manage that data, organize it, and make it shareable to people. Whether it is research, whether it is applications, whether it is whatever is going to benefit the industry. Everybody understands that. In this case, the University of Illinois is going to be that place.”
And in that neutral place, the future of the construction industry is going to grow.