We are proud to be a Laulima Sponsor and a Maui-grown company.
Vision Beyond the Horizon
Developing and bringing to market capabilities that are relevant to our national defense and security.
Supporting the TacSRT Mission Execution
Partnering with the U.S. Space Force Joint Commercial Operations (JCO) Cell to deliver rapid, reliable mission support.
Join us for our technical presentations and see how we’re shaping the future of space domain awareness.
The Space Force is actively engaged in addressing the complex operational landscape within Space Domain Awareness and Space Control (SDA/SC). Prior SDA/SC initiatives have encountered challenges in transitioning capabilities to operational effectiveness, often stemming from limitations in Testing and Evaluation (T&E) processes and a need for more comprehensive mission effectiveness analysis. Verification and validation of intricate SDA/SC systems requires testbeds capable of exercising the full spectrum of operational scenarios. Traditionally, system testing methods have relied on simplified models or limited datasets to verify functional performance. While valuable, these approaches may not provide the depth of evaluation required for the assessment of complex systems at the mission scope. To address these limitations, this paper introduces the Decisive Mission Analysis Capability (DMAC), a physics-based T&E framework designed to enhance the development and evaluation of SDA/SC capabilities.
A fundamental feature of DMAC is its modular software architecture, designed for plug-and-play integration of algorithms and modules from various providers. The framework is structured into abstract components representative of typical SDA/SC systems, including (but not limited to) image processing, orbit determination, orbital threat screening, and sensor tasker/scheduler modules. Importantly, the DMAC test harness supports both component-level and end-to-end system testing, aiming to establish a comprehensive and repeatable test environment. This approach facilitates rigorous evaluation of SDA/SC systems, enhancing their reliability and effectiveness, and promoting the development and validation of advanced algorithms and methodologies.
The ultimate objective of DMAC is to facilitate the quantifiable assessment of system performance and mission effectiveness. This is achieved through the two-tiered testing approach, encompassing both component-level and mission-level tests. Component testing allows for the isolation and detailed performance analysis of individual system modules, leveraging both real and simulated benchmark datasets. Examples of component test metrics include target precision and recall of optical imagery processing modules, or orbit accuracy evaluation from noisy, short-arc observation tracklets in orbit determination modules. Testing at the component level enables the assessment of functional correctness and performance against known ground truth in benchmark datasets. A key feature of the manuscript will be a detailed walkthrough of the component-level test metrics and methods, with the goal of establishing standard test measures of core SDA system components.
Alternatively, the mission-level testing capability of DMAC integrates all modules into a cohesive system under test, enabling system evaluation under simulated operational conditions. This test approach enables higher level analysis about system behaviors; specifically, whether or not the system meets mission requirements. In mission-level testing, full scenarios are executed with simulated sensor agents. Simulated sensor agents, driven by a full physics simulation backend, allow for a dynamic closed-loop workflow consistent with real-world operations. The full physics simulation is configurable to simulate diverse scenarios, with various space object populations, sensor locations, and orbital events. This simulator models orbital dynamics and sensor physics (e.g., optical, radar) at a high fidelity, generating realistic sensor agent data, which is in turn processed by the system under test in a closed loop until the scenario terminates.
The primary use-case of DMAC is to evaluate a system’s capability to acquire and maintain custody of orbital threats. In accordance with the problem statements of the SDA TAP Lab, two pertinent adversary kill chains in the current space domain are GEO Direct Ascent and Co-orbital anti-satellite weapons. The exquisite mission-level testing capability of the DMAC framework enables critical evaluation of a full integrated system in scenarios for which real data is not available, such as the above adversary kill chains. Further, the modularity of the DMAC framework allows swapping of system components to evaluate differences in mission effectiveness. A major focus of the manuscript results will be statistical analysis of a full SDA/SC system against scenarios involving GEO Direct Ascent and Co-orbital anti-satellite weapons.
Initial applications of DMAC have already demonstrated its potential for identifying system deficiencies. Recent testing of an examplar system revealed a weakness in the orbit determination component. When tracking an uncorrelated object on a Geostationary Transfer Orbit near perigee, the system under test was unable to maintain custody of the target after several minutes due to the inaccuracy of the orbit solution. The result prompted subsequent modifications to the relevant orbit determination components, enhancing the system’s future mission effectiveness. Similar analysis will be presented in the manuscript, extending to all system components.
In summary, DMAC provides a robust T&E tool for advancing the development and validation of SDA/SC systems. Its modular design and full physics simulation capability enables rigorous testing of both individual components and integrated systems, enhancing their reliability and effectiveness. Future work will focus on expanding support for a wider range of sensor types and operational scenarios, pushing for broader adoption within the SDA community.
Co-Authors: Alexander Cabello, EO Solutions; Virginia Wright, AFRL; Miguel Rodriguez, AFRL
The growing interest in Cislunar space has accompanied a surge in missions to the Moon and beyond by government, scientific, and commercial stakeholders alike. Although the motivations for their interests differ, the result of an increased presence in this regime remains. During the early days of spaceflight, a poignant lack of consideration was initially given to the ramifications of rapid growth in the space domain when it came to Low Earth Orbit (LEO). The vastness of space, combined with the extreme costs required to achieve orbit at the time, made it unnecessary to closely consider the positions of satellites, their relative proximities, or operational intentions. This mindset has dramatically shifted since then with the explosion of satellites operating from LEO to Geostationary Earth Orbit (GEO). As a result, the field of Space Domain Awareness (SDA) arose to tackle these new challenges.
The same SDA considerations for orbits solely around Earth must now be taken into account for those in the Cislunar domain. However, the specific challenges for Cislunar space differ from those of near-Earth orbiting objects. To maintain SDA, the requisite search volume of the Cislunar regime is over 1000 times greater than that of the GEO regime. Additionally, a majority of the sensors used for consistently monitoring Resident Space Objects (RSOs) are only designed for operational ranges up to GEO, severely limiting the resources available to be proactive about SDA in the Cislunar domain. Lastly, the dynamics of the region allow the existence of trajectories from periodic orbits that require very little fuel expenditure to go from the farthest regions around Lagrange points back to Earth.
Many proposed solutions for maintaining Cislunar SDA from the current literature involve placing space-based Cislunar assets on periodic trajectories where a significant variety of periodic orbits exist and possess great advantages due to their diverse geometries. However, building satellites with the correct capabilities, getting them to these exotic orbits, and maintaining the orbits are an extreme investment of time and money that may not be practical in the near-term. Repurposing currently operational infrastructure has the potential to begin tackling the problem immediately and address the growing SDA concerns in Cislunar space.
Poincaré Maps (PMs) offer a solution to the volume search problem that the Cislunar region raises. PMs are mathematical surfaces placed within a physical representation of space corresponding to a discrete dynamical system. Using PMs, surfaces can be placed in strategic locations to chart where bodies subject to the defining system dynamics cross the map surface. In the context of this paper, a PM is placed at a spherical distance of 4X GEO centered around the Earth. The choice of 4X GEO as the location of the PM is motivated by the strategic implications of the aforementioned free return trajectories to Earth and the nearby orbital regimes of interest. By enabling search of a sphere at a distance 4 times that of GEO, this gives decision makers the lead-time required to assess threats and react accordingly in a decisive manner, and by focusing on a subset of the Cislunar volume to search, the daunting task of maintaining Cislunar SDA becomes much more tenable.
Utilizing the Circular Restricted Three-Body Problem (CR3BP) to represent the dynamics and all 272,008 periodic orbits for the Earth-Moon system contained within the JPL Three-Body Periodic Orbit catalog, a reachability analysis is conducted. 25 equally spaced points in time around each periodic orbit are given small delta V’s of 1-10 m/s and one positional translation, typically used to calculate manifold trajectories, and are propagated for one year. The points where each trajectory crosses the 4X GEO PM are recorded. A representative range of the total number of crossing points, defined by a +/- 36 degrees latitude spherical segment from the Earth-Moon plane, in addition to the first boundary crossings are analyzed to highlight the underlying patterns the sum of these transfers from Cislunar loitering trajectories display on their near-Earth fly-bys.
Using the total of these crossing points, the complete 4 pi steradian search space is reduced by up to 85.3% to a mere 0.587 pi steradians. Additionally, this search space of the unreduced crossings covers 100% of the span of the GEO belt that is either currently monitored or desired to be monitored. In other words, at the bare minimum, accomplishing the goal of total persistent monitoring of GEO would also succeed at monitoring a large swath of crossing points from possible loitering locations in Cislunar space.
Utilizing these crossing points, studies will be conducted using the optical sensors employed within the Space Surveillance Network (SSN) to determine the performance of these sensors working collaboratively to search the 4X GEO PM crossing region for all crossing points, the 4X GEO PM crossing region for the first crossings, and finally compare them to the performance of the unreduced search space.
In addition, a supplemental study is conducted in which a typical catalog maintenance methodology for searching the currently desired regions of the GEO belt is employed and the 4X GEO PM first crossing points that are captured within the background of that search are recorded. From this simulation, this study demonstrates the possibility of maintaining Cislunar SDA through current GEO catalog maintenance operations.
Adoption of monitoring any of these reduced-area search spaces by the United States Space Force (USSF) and other partners will provide the initial steps needed towards building up the required capabilities for maintaining Cislunar SDA.
Additionally, as shown by the practical application towards current operations, its adoption and use will accomplish the goal of tactical persistent GEO monitoring that is currently desired.
Co-Author: Kevin Schroeder, Virginia Tech National Security Institute
In the domain of Space Domain Awareness (SDA), one of the key challenges is the variability in data quality and characteristics across different sensors. Changes in hardware, the use of multiple sensor types, and the gradual degradation of existing instruments can all lead to differences in noise profiles, sensitivity, and resolution. These variations can cause detection models to underperform, resulting in false positives or missed detections, particularly for small or dim objects such as faint satellites or space debris. Traditional solutions involve retraining detection models on newly annotated datasets for each sensor, but this requires significant manual effort, expert annotation, and time, making it impractical for large-scale or time-sensitive operations.
In this paper, we present a novel approach for sensor-specific model calibration through automated target injection. Our method leverages high-confidence detections from a heuristic annotator, segments and masks these targets, and re-injects varied versions into verified blank frames from the target sensor. This process generates synthetic training data that reflects the sensor’s native background, noise, and optical characteristics, improving the model’s sensitivity to faint objects while preserving a low false positive rate.
We further integrate this process into retraining pipeline. A baseline RetinaNet model is first trained on high-quality industrial-grade imagery, and then fine-tuned using the created dataset from the target sensor. By utilizing a pre-existing model from another sensor the adaptation process is efficient, reducing both computational cost and training time. This makes the method well-suited for rapid recalibration when switching to a new sensor, upgrading equipment, or compensating for changes in operating conditions.
Our approach eliminates the need for extensive manual annotation for each new sensor, enabling scalable and repeatable calibration across heterogeneous systems. This method significantly improves detection performance of pretrained models on new sensors, reduces false negatives, and maintains stability across varied sensor types, ensuring that SDA systems remain accurate and reliable in dynamic operational environments.
Co-Authors: David Chaparro, EO Solutions; Taylor Phan, EO Solutions; Zach Gazak, SSC/SZBA
Space Domain Awareness (SDA) systems require operators to manage increasingly complex sensor networks through specialized scripting languages and legacy interfaces. While automation has advanced sensor scheduling and data fusion capabilities, the human-computer interface remains a critical bottleneck that limits operational efficiency. We present an on-premises natural language processing (NLP) system that leverages a large language model (LLM) to enable natural language command and control of the MACHINA (Mission-driven Autonomous Collaborative Heterogeneous Intelligent Network Architecture) framework for space operations. Our approach uses a state-based graph architecture using a 24B parameter open-source model, supporting three core capabilities: autonomous objective definition that translates natural language into structured commands, sensor-target visibility assessment, and preliminary system performance monitoring. Through trajectory-based evaluation, we demonstrate that our agent achieves high task completion rates (>90%) while maintaining process fidelity and error recovery mechanisms. Our analysis shows that reliable agent-based control emerges with larger models (24B+) compared to simpler structured output tasks achievable with 7B models. Our findings suggest that natural language processing systems are able to replace specialized scripting languages and dedicated visual interfaces for some tasks. More broadly, our work validates a template for deploying LLM-based agents in defense environments with potential to enhance operational workflows while preserving human oversight.
Co-Authors: Alexander Cabello, EO Solutions; Garrett Fitzgerald, EO Solutions; Zach Gazak, SSC/SZBA; Justin Fletcher, USSF SSC/BCB
Book a meeting with us during AMOS week – let’s talk collaboration.
If you are determined, clever, and authentic, and if you are looking for a place where your work makes a difference, EO Solutions is the place for you. Join us and maximize your potential while contributing to the national defense and security of our country.
