Aerospace · Stochastic · Robotics · Exploration
We enable aerospace systems to operate autonomously — safely, intelligently, and collaboratively — in uncertain and extreme environments. Our work fuses rigorous mathematical foundations with hardware validation in the lab and the field.
// More on the lab channel ↗
An agent tracks the optimal path while a bounded uncertainty ellipse grows and shrinks; sampled trajectories stay inside a corridor that keeps the obstacle clear with high probability.
Three observer agents follow stable passive relative orbits around a target, sharing information to jointly inspect it under fuel and communication constraints.
Autonomous rovers traverse uncertain lunar terrain, leaving tracks as they map the surface and search for resources ahead of crewed missions.
At the heart of our work is one question: how can robotic systems make intelligent decisions in the face of uncertainty, while ensuring safety, robustness, and mission success? We formulate and solve information-cost stochastic optimal control problems, drawing on nonlinear control, distributional robustness, convex optimization, and probability theory.
We design and prototype multi-modal robots that traverse highly variable environments — planetary caves, ice crevasses, steep slopes. Inspired by platforms like EELS, they combine wheeled, legged, and slithering locomotion with onboard autonomy: dynamics modeling, real-time control adaptation, and perception-informed motion planning.
We develop scalable methods for perception-driven planning and control under bounded and unbounded uncertainty, using distributionally robust optimization, uncertainty quantification, and probabilistic inference.
We study how heterogeneous teams of robots — rovers, drones, and free-flyers — reason jointly, share information, and act cooperatively under resource, communication, and safety constraints: distributed task allocation, formation planning, and multi-agent motion planning with limited observability.
We enable autonomous in-space assembly and reconfiguration of large structures such as antennas and trusses — multi-spacecraft coordination, contact dynamics, and uncertainty-aware manipulation for next-generation orbital construction.
Assistant Professor · Daniel Guggenheim School of Aerospace Engineering
Dr. Nakka leads the AstroX Lab at Georgia Tech. His long-term vision is to enable trustworthy, adaptive, and resilient robotic systems that serve as partners in exploration, science, and infrastructure development in the most challenging environments known to humanity.