System-level autonomous decision-making for earth observation satellite systems

Autonomous operational strategies for Earth Observation systems are called to address the challenging task of coordinating multiple sensing assets to deliver very short revisit times and near-real-time access to data. Contributing to the on-going exploration of autonomous operations, this paper presents one particular solution based upon knowledge exchange (through Inter-Satellite Links) and spacecraft-level mission planning and resource allocation. The goals of such autonomous DSS is the constant monitoring of the Earth (i.e. global coverage) that collectively minimizes revisit times. The objectives of this work are twofold: to explore the suitability of decentralized, on-board decision-making for generic EO systems, and to propose a characterization framework to evaluate the goodness of autonomous operations. The application of a simulation-based design exploration is detailed in this paper to show the impact that some design parameters (structural and functional) present upon system behavior. The results show performance trends under different operational constraints, including energy or limited ISL capacity. Finally, this work also applies the autonomous framework to the PlanetScope constellation, highlighting its suitability for a large-scale DSS encompassing 117 nano-satellite units.Postprint (published version