Comenius Scholar Highlight: Reva Paul '27

The Comenius Independent Study Program offers Merle-Smith Campus sophomores, juniors, and seniors the opportunity to pursue their interests through self-directed research and creative work, showcasing critical thinking, curiosity, and a passion for learning within the Moravian Academy community.

Through her Comenius Project, Exploring the Applications of Neural Networks in Satellite Motion and Space Systems, Reva Paul '27 investigated whether artificial intelligence could provide a software-based alternative to expensive onboard satellite accelerometers. These instruments measure nongravitational forces, such as atmospheric drag and solar radiation pressure, that can affect a satellite's orbit and are critical for accurate orbit modeling and collision avoidance.

Using data from the GRACE Follow-On (GRACE-FO) satellite mission, Reva worked with neural network models to estimate atmospheric drag from GPS data and compared the models’ outputs to onboard accelerometer measurements to evaluate accuracy. She found that a physics-informed neural network achieved approximately 97% accuracy in estimating atmospheric drag, demonstrating the value of combining physical principles with machine learning approaches.

Through this project, Reva explored the intersection of artificial intelligence, physics, and aerospace engineering while examining how machine learning can be applied to real-world challenges in satellite operations. Her findings suggest that software-based approaches could one day help support Low Earth Orbit satellites that do not carry onboard accelerometers.

Building on this work, Reva was recently selected as an on-site NASA SEES Intern, a NASA-funded research program hosted by The University of Texas at Austin's Center for Space Research. This summer, she will continue researching similar topics through a new project in Earth and space science.