I'm Krishna Muvva, and I build autonomous flight across multirotor, fixed-wing, VTOL, and flapping-wing platforms, teaching them to see, decide, and navigate on their own. Right now, that means flying them into forest so dense that few have ever reached it, on a NASA-funded project at the Nevada Center for Applied Research, in collaboration with Goddard Space Flight Center.
"DLCL: Deep Learning-enabled Cooperative Localization for a UAV-UGV Team" accepted at the International Journal of Computer Vision.
Led a hands-on professional development session: how to build your own custom UAV and make it autonomous.
Joined the Nevada Center for Applied Research to work on a NASA-funded project developing wildfire-prevention UAVs, in collaboration with NASA Goddard.
UNL's highest recognition for doctoral research, for "Learn to Fly."
One of AIAA's highest graduate honors, awarded to a single graduate student worldwide for excellence in aerospace research. Featured in Aerospace America.
UNL's highest recognition for doctoral research, awarded for my dissertation "Learn to Fly: Enabling Deep Learning based Perception & Control in Aerial Robotics."
At the Nevada Center for Applied Research, on a NASA-funded project with Goddard Space Flight Center, I'm building UAVs that fly themselves where GNSS can't reach. SLAM-based navigation fuses LiDAR, cameras, and IMU/GNSS so the aircraft localizes, maps, and plans its own path in real time, dodging trunks and branches through forest so dense that parts of it are unreached, unmapped, unknown.
Deep RL for adaptive perception (Twin Delayed DDPG), obstacle-aware control, cooperative UAV–UGV localization, and reliable AI through control/perception co-design.
Physics-informed neural networks for fixed-wing navigation in turbulence, plus an open ROS2 bridge for X-Plane that brings full autonomy stacks to fixed-wing simulation.
Transitional flight for missions that need both: vertical launch from confined field sites and forest clearings, then wing-borne cruise for the range and endurance that long surveys demand.
Neuromorphic perception and control for flapping-wing UAVs: event-driven sensing matched to wingbeat dynamics for agile flight in dynamic environments.
Talks, flight tests, and drones doing drone things. See the fleet in the air on my YouTube channel.
I'm Krishna Muvva, and I build drones that fly themselves. My research brings together perception and control, teaching aircraft to see and understand the world and to act on what they see, so they can fly in the places that defeat ordinary autopilots: turbulent air, dense forest, and skies with no GPS signal to lean on.
I work across nearly every kind of flying machine, from multirotors and fixed-wing aircraft to VTOLs and flapping-wing robots. Right now, at the Nevada Center for Applied Research, I'm on a NASA-funded project building UAVs that navigate deep beneath forest canopies, where GPS can't reach, to collect sub-canopy data in support of forest management and wildfire prevention.
I earned my PhD in Computer Science at the University of Nebraska-Lincoln, where my dissertation, "Learn to Fly," received the Folsom Distinguished Dissertation Award. My work has also been recognized with the AIAA Orville and Wilbur Wright Graduate Award, given to a single graduate student worldwide for excellence in aerospace research.
Developing autonomous UAV systems for wildfire prevention under a NASA-funded project in collaboration with NASA Goddard Space Flight Center.
Aerial robotics across flapping-wing, multirotor, and fixed-wing platforms.
Perception/control co-design for small uncrewed aircraft systems; instructor for UAV robotics courses.
Learn to Fly: Enabling Deep Learning based Perception & Control in Aerial Robotics. ★ FOLSOM DISTINGUISHED DISSERTATION AWARD
Advisors: Dr. Marilyn Wolf & Dr. Santosh Pitla.
Thesis: FARMNAV-UAV, a field-ready autonomous UAV for trajectory tracking in agriculture. Advisors: Dr. Santosh Pitla & Dr. Marilyn Wolf.
AI, robotics, and augmented reality. Advisors: Dr. Edward Swan, Dr. Yang Zhao & Dr. Song Zhang.