Digital twins—virtual representations of real-world objects or systems such as a jet engine or power plant—are powerful computational tools for enabling engineers to better design, analyze, control and optimize complex systems.
Unlike traditional modeling and simulation approaches, there is a back-and-forth interplay between a physical system and its digital twin. “So, the data collected from the physical system is used to improve the computational models in the digital twin. Then, based on that data, the digital twin informs the analysis, control and perhaps maintenance of the real-world system,” says Harsh Apurva Sharma, who joined the UW-Madison Department of Mechanical Engineering as an assistant professor in August 2025.
Sharma’s research is focused on integrating underlying scientific principles and domain-specific knowledge with machine learning/AI techniques for design, analysis and control of complex and large-scale dynamical systems, with an emphasis on digital twins. In particular, he’s working to enable digital twins that are predictive and computationally efficient.
In other words, they’re more accurate and reliable.
“Essentially, I’m trying to embed the underlying physics of the problem into the design of the algorithm itself, which allows me to develop a data-driven model that respects physical laws and has predictive capability,” he says. “So, when the model is operating and encounters a situation that it might not have seen in the training data, it can still extrapolate to correctly handle that new problem.”
In his research, Sharma works on a wide variety of computational methods, ranging from traditional modeling and simulation approaches to recent developments in data-driven modeling and scientific machine learning techniques. His research has potential applications in diverse areas, including soft robotics, structural dynamics, astrodynamics and computational physics.
Sharma comes to UW-Madison from the University of California, San Diego, where he was a postdoctoral research scholar in the Department of Mechanical and Aerospace Engineering. He received a dual degree (BS and MS) in mechanical engineering from the Indian Institute of Technology-Bombay and a master’s degree in mathematics from Virginia Tech. Sharma earned his PhD in aerospace engineering from Virginia Tech in 2020.
Sharma says he was drawn to the UW-Madison Department of Mechanical Engineering because of its research excellence in computational methods and data-enabled science as well as its many opportunities for collaboration.
“I always like seeing my computational methods being applied to challenging real-life problems, and I want to pursue collaborations with faculty working in areas such as soft robotics and energy systems,” he says.
Sharma was hired through the RISE-AI initiative at UW-Madison, which Chancellor Jennifer Mnookin launched to strategically hire faculty members, develop research infrastructure, improve interdisciplinary collaboration and increase educational opportunities in artificial intelligence. “It’s exciting that UW-Madison is putting a lot of resources into this area and creating synergy between researchers working on different aspects of AI,” he says. “It makes UW-Madison a very attractive place for me to build my research program.”