Sparse Information Orbit Estimation for Proliferated LEO

Rapidly expanding Low Earth Orbit (LEO) satellite constellations force traditional ground-based tracking methods to adapt as the current ground sensor networks can no longer provide a data rich tracking environment. Accurate tracking information is continually consumed by the Government and private sector to varying degrees of accuracy throughout satellites’ mission lifecycles to provide and utilize services like earth-imaging, weather imaging, situational awareness, telecom, media distribution, and internet service provision. In cooperation with the University of Colorado at Boulder, Braxton overcomes the sparse tracking data problem by utilizing modern statistical and machine learning techniques to maintain adequate orbit estimation for large numbers of spacecraft. A single orbit estimation algorithm is unsuitable for addressing a multitude of situations including CubeSat swarm launches, orbit colocation issues, and object break-up scenarios; therefore, our solution provides a plug and play architecture in which multiple orbit estimation techniques process an object’s tracking observations and the optimal estimation technique is automatically identified and critiqued against varying criteria such as accuracy, processing requirements, and solution maintainability.