Tiffany Phan, University of Texas Austin; Moriba Jah, University of Texas Austin
https://doi.org/10.64861/HTJV9492
Keywords: hybrid modeling, contextual analysis, situational awareness, threat assessment, decision support
Abstract:
Space Domain Awareness (SDA) traditionally relies on orbit determination, maneuver detection, and proximity analysis to understand the behavior of space objects. These tools provide precise information about where objects are and how they move. However, analysts evaluating unusual maneuvers or proximity operations must also consider contextual factors such as operator history, publicly-stated objectives, and known technological capabilities. Physical data alone does not fully answer the question of whether we should be concerned about a space object’s behavior.
This work presents a hybrid framework designed to combine orbital information with contextual indicators in a structured and transparent way. The approach treats the operational environment as two connected layers: a physical layer describing orbital motion and a contextual layer capturing relationships between actors, technologies, historical behavior, and open-source reporting. These layers remain separate but inform a common threat assessment process.
Threat is evaluated through three practical questions that mirror analyst reasoning: 1) Is the action physically feasible, 2) Does the operator have the capability to perform it, and 3) Is there evidence suggesting intent? Opportunity is derived from orbital geometry, including relative distance, drift patterns, and maneuver potential. Capability and intent are informed by contextual information such as prior proximity operations, documented on-orbit servicing systems, public statements, and historical behavior patterns.
Rather than forcing these indicators into rigid thresholds or overly precise probabilities, the framework assigns bounded plausibility levels that allow uncertainty to remain visible. This prevents the system from overstating confidence when evidence is incomplete or conflicting. The three components are then combined using a structured fusion method that accounts for interaction effects. For example, close physical proximity combined with demonstrated servicing capability may elevate concern even if intent is unclear, while strong rhetoric without physical feasibility produces limited escalation.
To demonstrate operational relevance, the framework is applied to a representative geosynchronous proximity operations scenario involving sustained co-orbital drift and repeated station-keeping near a high-value asset. Orbital data establishes whether approach is feasible, while contextual indicators such as operator history and known technologies provide additional perspective. The resulting threat assessment evolves gradually as geometry tightens and contextual information accumulates, avoiding abrupt changes often seen in rule-based systems.
A key design goal is stability and transparency. The system is tested to ensure that small changes in data do not produce disproportionate changes in threat assessment. Structural analysis of contextual relationships helps explain why the system arrives at its conclusions, supporting analyst review rather than replacing it.
The result is a practical and interpretable framework that integrates orbital feasibility and contextual information without oversimplifying either domain. By combining physics and context in a structured manner, this approach supports more informed and defensible decision-making in unclassified SDA environments.
Date of Conference: September 15-18, 2026
Track: Space Domain Awareness