Wanjiku Chebet Kanjumba, University of Florida; Norman Fitz-Coy, University of Florida
https://doi.org/10.64861/DETE4249
Keywords: Space Object Identification, Docking Interfaces, HARMONIA, Structural Inference, Space Domain Awareness, Bayesian Sequential Inference, Graph-Based Feature Matching, Non-Cooperative Objects, Cross-Modal Sensing
Abstract:
Characterizing non-cooperative or tumbling space objects remains a persistent challenge for space domain awareness, particularly for defunct satellites and debris objects lacking active telemetry. This paper presents HARMONIA-ID, the identification module of the Harmonized Adapter for Robust Modular Operations and Interoperability Applications (HARMONIA) program, for inferring object identity through the detection of standardized mechanical docking interfaces. Focusing primarily on the Rendezvous-Agnostic Fixture and Transfer Interface (RAFTI) standard, the approach treats geometric and material signatures of the docking collar as a structural fingerprint exploitable by passive sensors. An analytical sensitivity model predicts that synthetic aperture (SA) imaging combined with sparse laser ranging returns is sufficient to discriminate among three representative spacecraft size classes (small, ~85-110 kg; medium, ~450-520 kg; large, ~2100-2450 kg) sharing the same RAFTI port, using contextual features such as port-to-body aspect ratios and local curvature gradients. A graph-based feature matching architecture, described analytically, correlates observed partial views against parametric models of RAFTI-equipped vehicles. Topological constraints inherent to the interface geometry reduce identification ambiguity compared to body-shape-only methods under equivalent observation conditions. The framework is extended beyond RAFTI to encompass legacy docking interface analogues, including separation ring signatures found on thousands of defunct LEO and GEO objects, substantially expanding the addressable on-orbit population. The framework requires no telemetry or cooperative signals, relying solely on structural priors derived from interface geometry. This work establishes that interface standardization, often viewed as a vulnerability, can enhance space situational awareness when leveraged through cross-modal inference.
Date of Conference: September 15-18, 2026
Track: Satellite Characterization