Prof. Priyadarshini Panda Joins Effort to Study AI Technologies

Tasked with advancing the robustness of neural systems, Priyadarshini Panda, assistant professor of electrical engineering, has joined a DARPA-funded consortium that is taking on challenges in microelectronic technologies.

Panda began her two-year term Jan. 1, 2021 with the Center for Brain-inspired Computing Enabling Autonomous Intelligence (C-BRIC), a multi-university center led by Purdue University that focuses on key advances for cognitive computing. It’s part of a project funded by the Semiconductor Research Corp. (SRC) and Defense Advanced Research Projects Agency (DARPA) through its Joint University Microelectronics Program (JUMP). The project includes faculty and students from 10 other universities, including Purdue, MIT, Princeton, UIUC, and PennState among others.

By assembling a team of leading researchers from the fields of machine learning, computational neuroscience, theoretical computer science, neuromorphic hardware, distributed computing, robotics and autonomous systems, part of C-BRIC’s aim is to enable a new generation of autonomous intelligent systems, such as self-flying drones and robotic assistants. Panda, whose participation is funded with a $500,000 grant, was brought on for her work on increasing the robustness of neural network or artificial intelligence (AI) systems - specifically, making them reliable enough to use them for more autonomous systems. Despite the many advances in the field, Panda notes that there are still several questions that limit their safe and reliable deployment in real-world applications.

“Most of the time, these neural networks or AI applications are doing a lot of predictions, but we don’t know if we can trust them,” Panda said. It’s a particularly critical issue now, as AI is becoming increasingly ubiquitous and used for everything from medical applications to self-driving cars - areas where an unreliable system could have catastrophic results. 

Her work with C-BRIC is part of the effort to better understand how to employ AI in drones, robots and other autonomous applications and how to give them the ability to adapt and think on their own. “If you’re giving AI that kind of power, you also have to start thinking about whether we can actually rely on these predictions,” she said.

Panda's participation will involve collaborating with other C-BRIC researchers, and she’ll report back regularly to the C-BRIC leadership. Panda’s membership with C-BRIC is also valuable for her students, who will have opportunities to present their research and interact with others in the organization.