AI for safe and resilient space operations

Artificial intelligence is transforming space operators' situational awareness, enabling decisive actions to protect missions, people and progress on Earth

Three points to remember 

1. Low Earth orbit (LEO) is growing increasingly crowded with resident space objects such as satellites, rocket bodies and associated debris from past missions, making protection of the satellites that power daily life and national security vitally important.  
2. Traditional analytics are struggling to keep pace with processing data on the increasing number of space objects, thus requiring AI-driven solutions that provide sense-making and rapid response.
3. Leidos and Kayhan Space, using AI, are providing timely insights from large volumes of raw observations from LEO to secure space and safeguard U.S. assets and operations. 

Space was once thought of as limitless. Today, the band known as low Earth orbit (LEO) is growing increasingly congested with human-made resident space objects (RSOs), such as satellites, rocket bodies and debris from past missions. From navigation and weather forecasting to military communications and disaster response, many parts of modern life and national security depend on satellites in space.

As more nations and private companies launch capabilities into orbit, the risk of satellite collision or interference grows. Satellites or other space debris not belonging to the U.S. can move unpredictably or even act with hostile intent. If a U.S. defense satellite or the Global Positioning System (GPS) is damaged or destroyed, it could endanger lives on Earth and trigger global tension.

Leidos is thus building space domain awareness (SDA) solutions that help operators understand what is in orbit and how objects are behaving.

The limits of traditional space tracking  

Radar systems, ground telescopes and catalogs of satellite orbits were once enough to keep track of a relatively small number of satellites. However, those tools were built for a simpler time.

“Traditional approaches were never designed for this volume or pace of change,” explained Matthew Blackston, a Leidos cyber and analytics principal investigator. “Our goal is to give operators a near-real-time and persistent understanding of space object behavior by boiling down large amounts of data into insight and action, helping the U.S. stay ahead in this critical domain.”

The challenge isn’t just knowing where a RSO is. Operators must understand what the object is, if its behavior changes unexpectedly and if that impacts other space assets. Even a short delay in determining that a newly identified object is a foreign satellite shadowing a U.S. asset could leave little time for an appropriate response. 

To maintain the nation’s advantage in space, we need strong SDA that comes from a diverse set of data.

Enter SILO: faster insights from LEO observations 

Leidos and Kayhan Space, a Broomfield, Colo.-based company specializing in space safety, co-developed two AI-powered applications, called SILO-ID and SILO-B&F, under Leidos' Space Insights from LEO Observations (SILO) effort for the Defense Advanced Research Projects Agency's (DARPA) Space-domain Wide Area Tracking and Characterization (Space-WATCH) program.

The SILO tools add interpretation to tracked objects. They combine observational data, astrodynamics, graph analytics and machine learning to create a living, operational view of what’s happening above Earth. Each application benefits from the large increase of observational coverage provided by Space-WATCH’s commercial data providers, but that increase brings a deluge of data to process.

SILO–ID identifies RSOs from observational data with a high degree of precision. From the large number of commercial sensors providing data for the Space-WATCH program, it pinpoints what’s in orbit for operators to confirm object identity, link observations over time and support safe mission operations.

SILO-ID works by turning sensor observations into tracks. It matches these tracks to known objects when possible, helping operators understand which objects are already cataloged.

When observations do not match anything in the catalog, SILO-ID identifies them as new objects. It groups related observations, estimates the new objects’ initial orbits and supports their identification. The process uses trusted physics-based methods, but faster machine learning approaches are now in development.

SILO-B&F, for behavior and feature, explains single and multi-object behavior in real time. For individual objects, the application detects when an object moves away from its expected path, identifies maneuvers and classifies the types of maneuvers performed. These updates keep object information current and support timely maneuver alerts.

For groups of objects, SILO-B&F looks for coordinated movements and unusual patterns. It tracks how satellites move relative to one another, establishes what normal behavior looks like and flags changes from that baseline. This allows operators to quickly spot coordinated maneuvers and understand their potential impact.

Why this matters for national security

The insights SILO provides are strategically vital because satellites underpin nearly every aspect of U.S. national security and civilian life. Satellites help with missile warning and reconnaissance to keep our military aware of threats, and they handle secure communications for coordination of defense and homeland security. GPS navigation supports everything from aircraft to ride-share apps. Weather and observation satellites forecast and monitor hurricanes, wildfires and severe storms.

“If we lose awareness in orbit, we lose the ability to protect assets that give our forces an edge,” said Seth Henshaw, senior director of autonomy and AI for several DARPA and Intelligence Advanced Research Projects Activity (IARPA) programs at Leidos. “From missile warning to secure communications, space superiority is national security. Additionally, loss of control over assets used daily by the nation’s population like GPS could disrupt traffic and navigation.”

Losing access to even one of these functions could have widespread consequences. A destroyed space asset could lead to rising conflict between nations. A GPS outage could ground flights, delay emergency responders and disable navigation apps, while disruptions to communications could interrupt financial transactions, affect internet connectivity and hamper military operations. Loss of weather data could leave populations vulnerable to natural disasters.

Securing the future of LEO 

Protecting LEO is more than defense; it’s about keeping people safe and connected. SILO builds on Leidos’ legacy of supporting critical space missions, showing how fast data processing and real-time surveillance can help strengthen the U.S. advantage in space.

SILO helps ensure the U.S. isn’t just observing space but also securing it.

------------

This research was developed with funding from the Defense Advanced Research Projects Agency (DARPA). The views, opinions and/or findings expressed are those of the author and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. government.

MORE ON LEIDOS' SPACE CAPABILITIES