Matthew Givens, Advanced Space LLC; Alexandre Cortiella, Advanced Space LLC; Amanda Marlow, Advanced Space LLC; Veronica Rankowicz, Advanced Space LLC; Andrew Koehler, Advanced Space LLC; Aaron Liao, Advanced Space LLC; Justin Spurbeck, Advanced Space LLC; Charles Cain, Advanced Space LLC; Patrick Miga, Advanced Space LLC
Keywords: Space Domain Awareness, Cislunar, Uncertainty Management, Uncertainty Propagation, Sensor Tasking, Custody
Abstract:
The number of objects in cislunar space, both cooperative and non-cooperative, continues to grow and is slated to grow significantly in the future. Due to the large volume of space involved, highly nonlinear dynamics, exclusion zones, and geometric information constraints of orbits distant from Earth, it is currently very difficult to track objects in cislunar space. Classical assumptions and techniques applicable to catalog maintenance, uncertainty management, and orbit determination are already stressed to their limits due to the earlier challenges mentioned. In the context of orbit determination, these factors can easily be leveraged by non-cooperative and adversarial agents to avoid custody and tracking, presenting new challenges and even threats to friendly assets and personnel. Therefore, algorithms which are robust to the challenges of tracking and navigation in cislunar space must be developed and matured.
Advanced Space is developing a new algorithmic pipeline to be used in a prototype uncertainty management suite with the purpose of maintaining custody of objects in cislunar space. The core of this work is focused on four key pillars; the first of these is algorithms focused on intelligent destination prediction; the second is algorithms and techniques for non-Gaussian, multi-modal uncertainty propagation; the third is to refine future observations with uncertainty-based sensor tasking; the fourth is feedback-driven cislunar catalog maintenance. These components form the backbone of the cislunar custody maintenance solution: the Cislunar Uncertainty, Surveillance, Tracking, and Orbit Determination Evaluator (CUSTOD-E) toolset.
Starting from some known hard data of a given space object, the destination prediction algorithms utilize any soft data available in the catalog to predict future qualitative behavior of the object. This subset of outcomes is then fed into a Particle Swarm Optimization (PSO) algorithm which enables the determination of possible outcomes that considers controlled inputs such as maneuvers and specific regions of interest that are more likely destinations for a given object. The algorithm quickly computes optimal and near-optimal transfers to each candidate outcome and also considers the nominal ballistic case. These solutions are then propagated, queried, and parameterized through the use of a Particle Gaussian Mixture Filter (PGMF) methodology. Given a sensor with specified capabilities and geometry constraints at some epoch, the PGMF output can be used to inform sensor tasking and, upon measurement generation, incorporate multiple model probabilities and negative information. Successful redetection of the object results in additional qualitative and quantitative information that can be fed back to the object profile and subsequently the catalog, and the process can then repeat.
The sensor tasking approach builds on several tools created by Advanced Space and bridges the gap between the analyst and the operator end-user. Informed by the most critical destinations and the availability of observing assets in cislunar space, existing ground-based sensors will be assessed on a variety of metrics on their ability to maintain cislunar custody. The placement of future space-based sensors will be assessed by examining existing gaps and developing metrics using the developed algorithms.
The expected outcomes of this work include details on theoretical foundations of these components, as well as analyses that show their efficacy, and additional commentary on results that did not work well. These will be presented and compared with existing approaches.
Date of Conference: September 16-19, 2025
Track: Cislunar SDA