Editor's Note: Originally published by the Oden Institute for Computational Engineering and Sciences, written by Tariq Wrensford.
Building a future where AI-driven solutions enhance construction, infrastructure and education has long been a passion behind Krishna Kumar's research. Kumar, an assistant professor in the Cockrell School of Engineering's Fariborz Maseeh Department of Civil, Architectural and Environmental Engineering, shared insights into his work, how artificial intelligence (AI) is shaping the new generation of engineers, and why outreach programs are key to sparking that interest in young minds.
Over the years, Kumar, who is also an affiliated faculty member at the Oden Institute for Computational Engineering and Sciences, has developed various outreach initiatives to engage young students in learning about AI. One of his projects is a coding camp in collaboration with the Austin Public Library (APL), where kids aged 7-12 learn how AI works, but more importantly, they discover how it can sometimes fail.
In the camps, students learn to build their own algorithms, such as convolutional neural networks (CNNs), which Kumar uses as a teaching tool. “I want students to understand the explainability part of AI,” he says, noting that AI models, while powerful, are not foolproof. A key example he gives involves the use of AI in identifying damaged buildings. In this scenario, the AI might misclassify an intact building as damaged simply because it spots a clear blue sky behind it—a mistake rooted in human-labeled data. “It sees a lot of blue sky, and it thinks the building has collapsed,” he explains. A reminder of how critical it is to ensure AI is thoroughly understood, not just used blindly.
Kumar’s AI-focused outreach does more than expose students to technology; it encourages creative thinking in civil engineering. He sees a future where construction processes are more automated, and cities are designed with both efficiency and people in mind.
“Civil engineering has always designed infrastructure without thinking about people. How can we change that?” Kumar ponders.
One of his AI initiatives examines how infrastructure differs between high- and low-income neighborhoods. When trying to map traffic signs or identify structures, his AI models revealed disparities in the accuracy of identifying objects in low-income areas, largely due to the complexity of those neighborhoods. “The data comes from places like Tesla and Waymo [known for their self-driving vehicles] which generally drive in high-income neighborhoods, not low-income areas,” Kumar explains. By familiarizing students with these challenges through AI models, Kumar hopes to instill in them the drive to address these societal inequalities.
Kumar is no stranger to the hurdles that come with pushing innovative programs. His coding camp with the local libraries took over a year and a half to establish. When asked about the bureaucratic red tape, Kumar admits that it was a challenge, especially when considering expanding to local schools such as Austin Independent School District (AISD). While he wants to reach more students, it requires more collaboration and support.
“I’ve done this for ten years, and I’m at the point where I think someone else should handle the bureaucracy,” he jokes. But the work continues, and his passion for developing AI curricula for teens and younger students remains undiminished.
This fall, Kumar is teaching a course called code city: program your world at the Austin Public Library, expected to be offered mid-November. In addition to this, Kumar is in the development stages of a program called AI for Teens, to teach students how to build AI algorithms that can identify structural damage in buildings. As part of his outreach efforts, Kumar put out a call for volunteers at an Oden Institute seminar last spring, which garnered support from several graduate assistants, including Nathan Tsao, Ismail Vurankaya, and And Yilmaz.
Looking forward, Kumar foresees a civil engineering field driven by automation, AI, and a more human-centered approach. He’s especially interested in using AI to help solve complex problems that arise when designing infrastructure for diverse populations. He hopes to inspire the next generation to think more holistically when designing solutions for the future.
With the foundation he’s laying through his teaching and outreach, Krishna Kumar is ensuring that the engineers of tomorrow are not only well-versed in the latest technologies but are also equipped with the knowledge to use them responsibly and creatively.
