Stefan Doucette, InTrack Radar Technologies, Inc.; Kaitlyn Raub, InTrack Radar Technologies, Inc.; W. J. Mandeville, InTrack Radar Technologies, Inc.; Tim McLaughlin, InTrack Radar Technologies, Inc.
https://doi.org/10.64861/PYOZ8336
Keywords: Uncorrelated tracks, UCTs, Pine Park Observatory, search,
Abstract:
Maintaining an accurate catalog of Earth-orbiting objects requires the weekly tracking of nearly 50,000 artificial satellites using ground-based optical and radar sensors. These objects range from large operational payloads, such as the International Space Station, to small debris fragments from past missions and on-orbit events [1]. Because orbit prediction models are approximate, some objects’ cataloged locations periodically fall outside expected uncertainty bounds and are effectively lost, increasing operational risk. When subsequently detected, these observations appear as uncorrelated tracks (UCTs), creating a challenging re-association problem.
InTrack Radar Technologies (IRT) has developed a prototype capability for the targeted recovery and resolution of unknown satellites not represented in the current public satellite catalog. Unlike traditional catalog maintenance approaches that focus primarily on re-observing known objects or performing broad-area survey scans (e.g., GEO belt searches), this capability replaces the manual, operator-driven search procedure previously used at IRT’s Pine Park Observatory (PPO) with an automated tasking architecture that autonomously prioritizes, searches for, and characterizes recently detected objects, often within hours of initial observation.
The system implements automated field-of-view search using SGP4-predicted states and real-time weather and image-quality feedback. Propagation-error analysis across GEO, MEO, HEO, LEO, and HAMR populations shows that along-track error consistently dominates the radial and cross-track components, and this result is used to justify a search geometry that expands only along the predicted direction of motion rather than resorting to inefficient wide-area scans. Closed-loop weather and star-visibility feedback allow the system to terminate an unproductive search early and reallocate observing time to the next highest-priority target.
Various search patterns are established and comparatively evaluated to determine performance sensitivity to orbital regime, error-growth characteristics, and time since element-set epoch. System performance is assessed using thousands of observations across hundreds of UCTs collected throughout 2025 and 2026.
This work demonstrates that an automated, priority-driven scheduling architecture, combined with an empirically justified along-track search geometry, enables more rapid and resource-efficient UCT resolution than manual, operator-driven search procedures. By integrating automated prioritization, along-track search design, and near-real-time environmental awareness, IRT’s capability provides a scalable framework for responsive catalog maintenance and lost-object recovery in increasingly congested orbital environments. Extending this framework toward per-object, covariance-informed search geometry is discussed as future work.
Date of Conference: September 15-18, 2026
Track: SDA Systems & Instrumentation