Spencer Folk has successfully defended his doctoral dissertation “Real Time Local Wind Inference for Robust Autonomous Navigation,” completed under the guidance of Vijay Kumar, Nemirovsky Family Dean and Professor, and Mark Yim, Asa Whitney Professor of Mechanical Engineering. Folk’s thesis introduces a new paradigm for enabling autonomous flight in complex, windy urban environments which allows aerial robots to understand and adapt to local wind conditions in real time, without prior environmental knowledge.
The path to this work was not straightforward. Beginning the Ph.D. in August 2019, Folk had only a short time on campus before the COVID-19 pandemic forced research and community life online. “It was really bad timing—I wasn’t yet established in a lab group,” he recalls. But things shifted when a NASA internship connected Folk with John Melton, who would become Folk’s mentor. “The internship helped me find an initial direction and really motivated me to keep going,” Folks says. Over the next two years, that seed grew into a dissertation bridging deep learning, fluid mechanics, optimal control, and real-world experimentation.
At its core, the thesis tackles a pressing challenge in urban air mobility and autonomous package delivery: how aerial robots can operate safely and efficiently despite highly variable and unpredictable wind flow fields. Instead of relying on precise environmental maps or distributed wind sensors, Folk demonstrates how robots can reason about their surroundings using only onboard LiDAR and sparse in situ wind measurements. The research develops a framework for predicting local wind fields and integrating this information into motion planning, resulting in improvements in both energy efficiency and collision avoidance. Simulated demonstrations across diverse urban wind environments quantify these gains, while sub-scale free-flight experiments in an open-air wind tunnel show that the algorithms run robustly in real time on embedded flight hardware. The work ultimately contributes a new way of thinking about localized wind inference—one that pushes autonomous aerial systems closer to safe, reliable deployment in complex environments.
Alongside the technical milestones, Folk’s Ph.D. years were marked by moments of community, creativity, and discovery. Some of the early pandemic period feels like a blur, but he fondly recalls taking classes and organizing social events for the Mechanical Engineering Graduate Association with Greg, Jake, and Parker. A spontaneous decision to enroll in a manufacturing course became another highlight: building a Stirling engine in the Precision Machining Laboratory served as both a hands-on escape from thesis pressure and a reminder of the joy of making. There was also the International Conference on Robotics and Automation (ICRA) 2025, where over 14 lab members presented research, shared an Airbnb, and celebrated a milestone conference together.
Images L to R: Folk (right) with Peter Szczesniak (left) and the stirling engine he built; Folk with his mentor, John Melton (left), after finishing their last round of experiments; Kumar Lab from ICRA 2025 in Atlanta
Most unforgettable, however, were the repeated trips to the NASA Ames Research Center to run Folk’s thesis experiments. “These trips were physically and mentally draining,” he admits. “But I had privileged access to run my research code in a world-class wind tunnel facility.” Standing in a space with such a storied history of aerodynamic testing left a lasting impression.
Looking ahead, Folk will begin a new role in March as an Applied AI/ML Researcher at Aurora Flight Sciences, where they will continue advancing autonomy, modeling, and intelligent aerial systems.
To future MEAM Ph.D. students, he offers: “Embrace collaboration and reach out early to people both within and beyond your field. And take a class or two simply because it sounds fun or pushes you into new territory.”
Folk extends deep gratitude to the people who shaped this journey, including but not limited to: “my advisors, Mark and Vijay, for their support throughout my Ph.D. I’m also incredibly grateful to my NASA mentor, John Melton, for his guidance and especially for supporting my experiments at NASA Ames. Finally, thank you to all the friends who created such a welcoming and safe environment during my time at Penn.”