Geometric-Based Distributed Autonomy Approach for Data Fusion

The increasing expanse of autonomy towards space and in particular, the leveraging of clustered satellite systems has become increasingly important in a spectrum of areas that include, but not limited to, situational awareness, repair, and reliability, to general guidance and navigation control. At the heart of space-based autonomy is the ability to extract from a given set of observations, the shape and pose of a given object(s) from which higher order situational tasks can occur (e.g., control and navigation). To do so, recent space architectural advancements have pursued the use of distributed clustered satellite systems such that one can couple multi-perspective observations to pursue a common task for increased mission assurance in contested environments. Classic terrestrial distributed control of such agents has provided the opportunity for increased reliability against faltering agents while also increasing algorithmic efficiency. However, whereas these terrestrial counterparts are often well-studied, the remote nature and limited resource environment of space presents challenging algorithmic problems some of which include low bandwidth, latency, to physical weight constraints and computational power. As such, this effort will seek a geometric-based distributed autonomy approach for data fusion with aims to develop flight ready software.