Dustin Burchett, MITRE Corporation
https://doi.org/10.64861/SOKR7447
Keywords: ORCHESTRA, Space Domain Awareness, Space Surveillance Network, sensor scheduling, sensor orchestration, linear programming, static mission planning, custody and catalog maintenance, MACHINA, Unified Data Library, machine-to-machine tasking, optimization
Abstract:
Approved for Public Release; Distribution Unlimited. Public Release Case Number 26-0434_x000D_
Modern Space Domain Awareness (SDA) requires global sensor network coherence while enabling responsive adaptation to dynamic threat environments. Legacy Space Surveillance Network (SSN) sensor tasking paradigms provide coarse guidance for routine missions such as custody and positional awareness, while dynamic missions are handled through manual, direct command. This paradigm forces sensors to self-schedule in isolation, requires human-intensive deconfliction, and cannot be systematically re-optimized as conditions change throughout the day, leading to inefficient use of sensing resources and degraded planned mission performance that undermines time-dominant threat response._x000D_
This paper presents ORCHESTRA, a mathematical framework for systematic SDA sensor orchestration focused on planned (static) mission coordination across heterogeneous sensors. ORCHESTRA is designed to complement near-real-time decentralized approaches such as Sensor Network Autonomous Resilient Extensible (SNARE), which transform sensor tasking into locally optimized, time-dominant decisions based on continuously updated state. In the paired paradigm, decentralized methods support unplanned (dynamic, reactive) missions that demand immediate sensor availability, while ORCHESTRA provides mathematically optimal planned coordination that increases baseline mission quality and preserves flexibility for responsive tasking._x000D_
ORCHESTRA formalizes a practical, machine-to-machine (M2M) compatible SDA ontology that links belief states to executable tasking. A belief state represents the set of representations estimating environment attributes. Objectives represent desired states and constraints of the universe, typically instantiated as objective-target pairs with explicit priorities and collection constraints. Objectives compile into intents, the atomic executable units that translate constraints into permissible actions. Schedules map intents to sensor-time slots, enabling direct translation into tasking commands. ORCHESTRA produces a Time-Ordered Task List (TOTL) that can be compiled into Unified Data Library (UDL) compliant collectRequests, supporting M2M tasking pathways used by modern SDA systems._x000D_
The core technical contribution is a two-level scheduling architecture based on linear programming. Long-Term (LT) scheduling assigns objective-target pairs to feasible sensor-time intervals over an operational horizon, enforcing priority dominance while encouraging desirable observation spacing and efficient resource usage. Short-Term (ST) scheduling refines interval assignments into feasible sensor-time bins, producing precise bin-level coordination while enabling resource minimization and operational efficiency. Feasibility matrices encode sensor-target-time constraints, including objective-specific collection requirements that can be captured as linear constraints. The framework is designed to prioritize high-value objectives when the problem is over-subscribed, ensuring that critical missions are never compromised for routine tasks._x000D_
ORCHESTRA is naturally compatible with the objective-intent-action decomposition used by the government-owned Mission-Driven Autonomous Collaborative Heterogeneous Intelligent Network Architecture (MACHINA) framework for autonomous SDA coordination. In particular, ORCHESTRA provides mathematically optimal planned coordination as a counterpart to dynamic, time-dominant decision making, and supports M2M orchestration through standardized tasking interfaces. The expected outcome is an enterprise-capable approach to planned SDA scheduling that reduces human-in-the-loop burden, improves systematic coordination across multiple SDA objective types and priorities, and preserves sensor capacity for responsive, dynamic tasking as conditions evolve._x000D_
The framework has been exercised in representative operational scenarios and integration-oriented testing with realistic SDA objectives and sensor architectures, demonstrating end-to-end generation of coordinated planned tasking products suitable for M2M interfaces. These evaluations indicate that systematic, priority-driven planned coordination can improve surveillance effectiveness and sensor utilization compared to non-coordinated operational approaches, while producing schedules fast enough to support operational replanning at machine speed._x000D_
_x000D_
The author’s affiliation with The MITRE Corporation is provided for identification purposes only, and is not intended to convey or imply MITRE’s concurrence with, or support for, the positions, opinions, or viewpoints expressed by the author._x000D_
©2026 The MITRE Corporation. ALL RIGHTS RESERVED.
Date of Conference: September 15-18, 2026
Track: Space Domain Awareness