Aditi Mallavarapu’s engineering education started as a small child, with regular visits to science centers and summer workshops.

“My parents fed my curiosity,” she said. “I would engage in open-ended exploration with building and tinkering projects related to electronics, robotics, biology, computer programming and geography.”

Now an assistant professor in the Department of Computer Science at NC State University, Mallavarapu investigates human-centric, data-driven methods that inform how learning takes place in open-ended environments.

“AI and data-driven computation has a lot of potential, but we will only be able to leverage this potential if AI is designed to mimic human ways of thinking and doing for the good of humanity,” she said.

This interview with Mallavarapu is part of an ongoing series in which we ask NC State faculty about their academic research, their motivations and their advice for students.

What is your area of expertise, and what motivated you to pursue it?

My area of expertise is in digital, interactive informal learning environments. When designed well, these environments mimic their real-world counterparts, offering learners safe and effective opportunities to explore the many different configurations.

I was first introduced to this research when I met my mentor, Leilah Lyons, during my last semester earning my master’s degree in computer science at the University of Illinois at Chicago. She was working on an urban planning game called EcoCollage, which really piqued my interest. My work with EcoCollage was regarded as the first to apply data-driven computational methods to open-ended learning environments. The fact that my engineering and computer science skills would have a direct impact with how people learn was very rewarding for me personally. This motivated me to come back to academia to pursue a Ph.D. after spending two years in industry.

What kinds of larger issues do you think your research could solve?

My work with open-ended learning environments supports training to address real-world complex systems problems and provides experts with the tools to simulate their decisions. My current projects deal with designing exhibits to foster awareness about antimicrobial resistance, food webs and safe urban planning decisions.

Training future generations the skills they need to address these issues, while engaging the current decision makers to mitigate the issues, has been a major motivation for these projects. Most social and global problems that we face today like climate change, disease progression, evolution and pollution are complex systems problems. These are characterized as having many interacting variables that operate at different scales of time, space and organization. They often lack one correct solution, making them open ended.

Creating an AI that could aid humans in understanding, addressing and managing these problems requires data from these situations. Learners engaged in these environments can contribute data that can be used for two purposes. The first is to train humans to work hand-in-hand alongside data-driven machines, contributing to the effective human-AI teaming. The second is to train an AI with these human-derived decisions, which was not possible before due to the lack of contextual human behavior data. This makes the AI more sensitive to human ways of thinking and doing, contributing to the design of human-centered systems and truly augmenting human decision-making processes for critical global and social problems.

Why is university-based research important?

University research brings together experts, advanced facilities and diverse student talent, enabling interdisciplinary solutions to complex problems. Many of the complex problems we face today require interdisciplinary approaches and cannot be solved by one single discipline or by a single person alone.

This fusion, where multiple experts can come together and combine each of their priorities with support from the university administration, looks different for every group and is critical to maintain for the sustained success of university-based research. A big part of university-based research is this collaboration of current experts and future experts, scientists and policymakers for the greater good of the society. I have particularly been able to work across many different problems due to the interdisciplinary collaborations and support from NC State.

Describe your research in five words or less.

AI for exploring complex systems.

What advice do you have for a student who wants to pursue similar research?

Select the problem you are personally motivated to solve, engage with involved humans to understand if the AI is generating useful information, and do not shy away from asking tough questions.

This post was originally published in College of Engineering News.