H. Liz Solera, Massachusetts Institute of Technology ; Jacob Zelko, University of Saskatchewan; Richard Linares, Massachusetts Institute of Technology
https://doi.org/10.64861/WTZJ1186
Keywords: Big Data Analytics, Satellite Behavior Characterization, Data Fusion, SDA System Interoperability, Data Sharing
Abstract:
Despite the growing emphasis on bridging proprietary space object catalogs, diverse surveillance phenomenologies, and distributed analytic pipelines, adequate data conventions do not exist for rigorously migrating advanced satellite behavioral constructs across organizational interfaces. The Consultative Committee for Space Data Systems (CCSDS) publishes architectural guidance and standardized formats for numerous space data products, including Orbital Data Messages (ODMs); however, ODMs and maneuvers alone do not provide a sophisticated enough representation of satellite behavior to meet the demands of modern Space Domain Awareness (SDA).
Absent a shared semantic layer, behavioral characterizations produced under one representation cannot be systematically translated into another, nor can the coherence of such migrations be verified, bottlenecking complex insights within the tool that produced them. Formalizing behavioral knowledge offers a tractable path forward by enabling specialized behavior constructs to be shared across systems and reused without being re-derived from raw data each time.
This work explores why, how, and to what extent such formalization is necessary and highlights the importance of maintaining clear logical separation between the satellite actor, its manifested behaviors, and the underlying observational processes to preserve data provenance and support analytic integrity.
The paper provides worked examples for how a compositional approach can translate between competing classification schemas, quantify their disagreement, and enrich behavior understanding independently of underlying data products. Ultimately, this work provides a foundational requirements blueprint for an incrementally adoptable SDA semantic model of satellite behavior, offering a path toward the portable and composable knowledge products necessary for machine-speed Space Domain Awareness.
Date of Conference: September 15-18, 2026
Track: Space Domain Awareness