Abstract:
Due to the rapidly-increasing complexity and scale of outer-space missions, incorporating autonomy into future spacecraft operations is necessary to ensure that they both “live long,” such as by avoiding collisions, and “prosper” by maximizing their task-specific objective function. In this talk, I present methods of incorporating verifiable autonomy into a pair of pertinent settings, namely, in spacecraft proximity operations and in sensor selection for Earth-observing constellations. Both settings characterize the spacecraft environment in terms of finite-state models that capture nondeterministic and stochastic uncertainties in order to develop algorithms for efficiently synthesizing verifiably safe strategies. In proximity operations, the resulting strategies are the optimal relative motion plans of the spacecraft, while in Earth-observing missions, they are the optimal attitude sequences to follow. I show that these approaches allow the efficient synthesis of optimal strategies for complex, safety-constrained missions that remain faithful to the low-level dynamics, energy constraints, and sensing capabilities of the spacecraft.
About the speaker:
Michael Hibbard is a Ph.D. student in the Department of Aerospace Engineering and Engineering Mechanics at the University of Texas where he works with Dr. Ufuk Topcu and Dr. Takashi Tanaka in the Center for Autonomy. He obtained his M.S. in aerospace engineering, also at the University of Texas, in 2020 and his B.S. in engineering mechanics and astronautics at the University of Wisconsin in 2018. His research primarily focuses on developing theory and algorithms for autonomous spacecraft operations, often drawing from elements of formal methods, optimization theory, and game theory.
