OrbitalGuard combines orbital data processing and propagation-based analysis methods to help users understand potential close approaches between objects in orbit. This page explains where our data comes from, how we process and propagate it, how we screen for conjunctions, and, just as importantly, the limitations you should keep in mind when reading the results. Our aim is to be transparent about our methods, so that the engineers and researchers who rely on them can judge the outputs with the right context rather than treating them as a black box.
OrbitalGuard works with publicly available satellite catalog information. Public catalogs track a large population of objects in orbit and describe their motion using orbital elements, most commonly distributed as two line element sets, often abbreviated as TLEs. Each set encodes the parameters needed to model an object's orbit relative to a reference time, including its shape, orientation, and position along the orbit.
The platform analyzes objects using these available orbital elements. The quality of any analysis depends directly on the quality of the source information. Orbital elements are derived from tracking observations, and their accuracy varies between objects depending on how frequently and how precisely each object is observed. A well tracked, actively maintained satellite may have more reliable elements than a small piece of debris observed less often.
Orbital datasets also change over time. Elements are updated as new observations arrive, older data becomes stale, and objects can enter or leave a catalog. Because of this, users should always consider the age and origin of the underlying data when interpreting OrbitalGuard outputs. The platform reflects the public information available to it, and it inherits both the strengths and the limitations of that information.
The orbital regime an object occupies also matters. Objects in low Earth orbit experience meaningful atmospheric drag and tend to have orbital elements that change more quickly, while objects in higher orbits are influenced by drag far less. This is part of why data age and object type both affect how much confidence to place in a given analysis, and why context around each object is as important as the raw numbers.
Orbit propagation is the process of estimating where an object will be in the future based on its current orbital elements. Starting from a known state at a reference time, propagation models the motion of an object to project its position and velocity forward in time.
Propagation is what allows OrbitalGuard to reason about future geometry rather than only a single snapshot. By estimating where objects may be located over time, the platform can look ahead to identify moments when two objects are expected to pass close to one another. The motion of an object in orbit is shaped not only by gravity but also by perturbing effects such as the flattening of the Earth and atmospheric drag in lower orbits, and these effects influence how an orbit evolves.
It is important to understand that propagation is an estimate, not a certainty. Predictions depend on the accuracy of the starting elements and on how well the model captures the real forces acting on an object. Atmospheric drag in particular is difficult to predict precisely, because it varies with space weather and with an object's shape and orientation. As a result, uncertainty grows the further you project into the future. A position estimated for the next few hours is generally more reliable than one estimated several days ahead. OrbitalGuard is built around this reality, and we encourage users to treat longer range predictions with appropriate caution.
Conjunction screening is the process of comparing objects to find potential close approaches. Once orbits are propagated across a time window, OrbitalGuard examines the relative positions of objects to identify moments when they are expected to come near one another.
When a pair of objects is projected to approach within a distance of interest, the platform can highlight that event, including when the closest approach is expected to occur and how close the objects are estimated to come. In the language of the field, these correspond to the time of closest approach and the estimated miss distance. Presenting screening this way helps surface the situations that may warrant a closer look, rather than requiring you to sift through an entire catalog by hand.
Screening typically works across a defined time window and focuses attention using a distance of interest, so that the objects worth reviewing rise to the surface while distant, unrelated objects do not add noise. Because the underlying positions are propagated estimates, a screened close approach describes an expected geometry based on current data, and it should be revisited as fresher data becomes available and as the moment of approach draws nearer.
Screening produces analytical insight, not an operational verdict. A highlighted close approach is an indication that a situation may be worth reviewing, based on the available data and modeling. It is not a statement that a collision will or will not occur, and it is not a decision. The purpose of screening is to focus attention, so that human expertise can be applied where it matters most.
Building on screening, OrbitalGuard presents information that helps you evaluate the orbital situations you care about. Rather than reducing everything to a single figure, the platform aims to give you the context needed to reason about risk for yourself.
This includes information about potential close approaches, such as their timing and estimated proximity. It includes the relationships between objects, so you can understand which objects are involved in a given approach. And it includes orbital risk indicators that help you compare situations and decide where to focus your attention.
The interpretation of this information remains with you. OrbitalGuard is designed to inform your assessment, not to make it. You and your team remain responsible for operational decisions, including any decisions about how to respond to a potential close approach. The platform provides analytical support for those decisions, drawing on public data and established methods, while leaving judgment where it belongs. For guidance on applying these insights within the platform, see our documentation guide, which walks through how to read and use the platform's outputs.
At a high level, OrbitalGuard follows a consistent workflow that turns raw orbital information into insights you can read. Each step is designed to be repeatable and transparent.
The platform gathers orbital information for tracked objects from available public sources.
Raw orbital elements are parsed and organized into a consistent, usable format, so that objects can be compared on a common basis.
The platform examines object positions and orbital characteristics, propagating orbits so that positions can be assessed across a chosen time window.
The results are translated into risk related insights, highlighting the close approaches and orbital relationships that may be worth your attention.
We believe an honest account of limitations is essential to a tool like this. None of this makes orbital analysis unhelpful, it is used across the space community precisely because it provides valuable awareness — that value is realized when results are interpreted with their limitations in mind.
OrbitalGuard depends on publicly available orbital information. Not every object in orbit is tracked in public catalogs, and coverage varies. The platform can only analyze what the available data describes.
Orbital elements are estimates derived from observations, and their accuracy differs from object to object. Some objects are tracked more frequently and precisely than others, and that variation carries directly into any analysis based on those elements.
Public data is refreshed on the schedule of its sources, not continuously. There can be a gap between the real state of an object and its most recent published elements, and two objects in a single analysis may be based on elements of different ages.
Propagation is inherently uncertain, and that uncertainty grows over time. Estimates of future position and close approach geometry should be understood as approximations with margins of error, not exact forecasts.
OrbitalGuard is designed to support awareness and analysis, and it works best as one input among several. The platform helps you see and reason about your orbital environment, but it does not replace the broader operational picture that experienced teams bring to their missions.
We encourage users to apply OrbitalGuard alongside mission expertise, professional judgment, and appropriate operational procedures. Where a situation carries significant consequences, the platform's outputs should be validated against other trusted sources and considered within your own operational framework. Used this way, OrbitalGuard adds a clear and accessible layer of orbital awareness to the decisions your team is already equipped to make.
OrbitalGuard was built to make orbital data more understandable for the teams working with it. If you would like to see how the platform can help you better understand the orbital environment around your satellites, we invite you to explore it further or to get in touch with us.