Gavin Hofer, Katalyst Space Technologies; Gabrielle Jones, Katalyst Space Technologies; Will Oldroyd, Katalyst Space Technologies
Keywords: Space Domain Awareness, Space Battle Management, Maneuver Detection, Uncorrelated Track Processing, Uncorrelated Track Resolution, Machine Learning, Post-Maneuver UCT Resolution, Object Identification
Abstract:
Katalyst has identified efficient and autonomous uncorrelated track (UCT) resolution workflow to be a key bottleneck in the ability to quickly identify large and potentially consequential maneuvers. A UCT can be resolved when it is re-associated to a previously known object track and the full state history of the object becomes available. However, current automated systems falter when they encounter maneuvers that break the kinematic constraints of observation correlation algorithms, leading to UCTs. Traditionally, observation association algorithms rely upon kinematic correlation algorithms, which work by propagating the orbital dynamics of catalog objects and comparing the propagated states with observations. These algorithms work by setting thresholds around the propagated orbit, and generally only consider the position and velocity of objects, leaving out additional information such as visual magnitude. This approach can fail when correlating across a maneuver because, based on state alone, once the deviation exceeds the thresholds it often fails to correlate with any objects in the catalog. This causes UCTs to drop into manual processes executed by analysts, resulting in hours of delays trying to resolve them. Additionally, there is not enough bandwidth to monitor all objects, so unknown threats, such as payloads mislabeled as rocket bodies, may be missed entirely when they maneuver.
To overcome these challenges, Katalyst has developed an automated post-maneuver UCT resolution workflow that integrates kinematic correlation with non-kinematic correlation using statistical models for both candidate filtering and to increase the confidence of correlations. To achieve this, the workflow incorporates statistical distributions of observable quantities generated from historical observations on objects in the catalog, which can be compared to the corresponding values in the UCT observations to adjust the confidence of the correlation. This workflow is agnostic to sensor type and can incorporate available observables associated with the UCT from electro-optical sensors. The Post-Maneuver UCT resolution workflow is composed of five main components. First, Stochastic Initial Orbit Determination (SIOD) creates a probability distribution for the object’s state based on the UCT observations. Second, Statistical Object Identification compares observable non-kinematic quantities of the UCT to distributions stored in a database for filtering and improved confidence determination downstream. Third, Candidate Filtering filters UCT-object candidate pairs based on both kinematic and non-kinematic quantities from the SIOD and Statistical Object Identification steps. Fourth, Probabilistic Kinematic Track Correlation finds the feasibility of correlation between the UCT and catalog object by propagating the probability density function (PDF) of the object forward and of the UCT backwards to find the maximum total probability of the product of the two PDFs at intermediate points. Lastly, Confidence Determination computes the correlation confidence for each candidate pair using a two-stage machine learning algorithm that considers the kinematic feasibility, the scores from statistical object identification, and the initial confidence (where initial confidence does not consider other possible candidates) of all other possible candidates for the UCT. Once the confidence for a candidate pair exceeds an upper threshold, messages are generated for the correlated object. This workflow for post-maneuver UCT resolution can correlate UCTs to objects with fewer observations compared to traditional methods.
Date of Conference: September 17-20, 2024
Track: Space Domain Awareness