The safety margin has become thinner. Low Earth orbit now depends on continual coordination.
Humanity has spent decades filling the skies above with satellites, each one a small claim on a shared commons that no single authority governs. A new research metric called the CRASH Clock now reveals that the safety margin in low Earth orbit — the time available to recover if the coordination systems keeping thousands of spacecraft apart were suddenly to fail — has collapsed from 164 days in 2018 to just 5.5 days in 2025. The cause is not negligence but arithmetic: megaconstellations have concentrated thousands of spacecraft into the same valuable orbital bands, tightening the system's dependence on continuous, flawless cooperation. What was once a forgiving environment has become infrastructure, and infrastructure, as history reminds us, is only as resilient as the coordination holding it together.
- A safety margin that once gave operators nearly six months to recover from a coordination failure has shrunk to less than a week — a change in scale that is difficult to overstate.
- Megaconstellations like Starlink have packed thousands of satellites into the same high-demand orbital shells, turning what were once sparse lanes into crowded corridors where close approaches are routine.
- The real danger is not a single collision but a cascade: one serious fragmentation event seeds debris that cannot maneuver, threatening other satellites and potentially rendering entire altitude bands unusable for years.
- Solar storms, cyberattacks, or software failures could simultaneously degrade the tracking and communication systems that make constant collision avoidance possible — and the window to recover has nearly closed.
- Researchers are offering the CRASH Clock not as a doomsday forecast but as a stress indicator, a way to make visible the growing fragility of an orbital environment that the world increasingly depends on.
In 2018, a model of low Earth orbit suggested that if every satellite suddenly lost the ability to dodge its neighbors — if tracking failed, communications went dark, and the entire coordination system collapsed — it would take 164 days before a catastrophic collision became likely. That was the buffer. By 2025, the same measure had fallen to 5.5 days.
The figure comes from a preprint paper by Sarah Thiele, Skye R. Heiland, Aaron C. Boley, and Samantha M. Lawler, who propose a metric they call the CRASH Clock — Collision Realization And Significant Harm. It is not a prediction of imminent disaster. It is a measure of how much slack remains in the orbital system if something goes broadly wrong. The distinction matters, but so does the direction of change.
The mechanism behind the collapse is straightforward. Satellites cluster in similar altitude bands because those orbits are most useful for communications and Earth observation. When megaconstellations like Starlink — and soon Amazon's Kuiper and several Chinese broadband systems — place thousands of spacecraft in those same shells, the number of close approaches multiplies. Each new satellite consumes a small portion of a shared safety margin that no single operator owns.
The debris dimension makes the stakes concrete. NASA estimates hundreds of thousands of marble-sized fragments already orbit Earth, alongside more than 100 million smaller objects. A collision is not an isolated event: the 2009 impact between Iridium 33 and the defunct Kosmos 2251 created a debris field that satellites still navigate around today. Fragments too small to track but large enough to destroy a spacecraft can remain aloft for years, and they cannot respond to a warning.
What the CRASH Clock ultimately measures is dependency. Low Earth orbit now functions because tracking systems, communication networks, and operators across many organizations keep working together, continuously, without failure. A solar storm distorting atmospheric drag predictions, a cyberattack disrupting communications, or a software fault cascading across multiple operators could each erode that coordination. The orbital environment has not broken. But the margin for error has quietly, and dramatically, narrowed.
In 2018, low Earth orbit still had months to spare. If every satellite suddenly lost the ability to dodge its neighbors—if operators went blind to where objects were headed, if communications failed, if the whole coordination system that keeps thousands of spacecraft from colliding simply stopped working—the model suggested it would take 164 days before a catastrophic impact became likely. That was the safety margin. That was the buffer between a functioning orbital system and a cascade of collisions that could render entire altitude bands unusable for years.
By 2025, that margin had collapsed to 5.5 days.
The measurement comes from a research paper titled "An Orbital House of Cards: Frequent Megaconstellation Close Conjunctions," authored by Sarah Thiele, Skye R. Heiland, Aaron C. Boley and Samantha M. Lawler. They propose a metric called the CRASH Clock—Collision Realization And Significant Harm—as a way to quantify how much operational slack remains in low Earth orbit if the system that keeps it functioning suddenly breaks. The paper is a preprint, not yet peer-reviewed, and it models a stressed scenario rather than predicting imminent disaster. But the comparison at its heart is difficult to ignore: the safety margin has tightened from months to less than a week.
The CRASH Clock is not a doomsday timer. It is a measure of environmental stress. Satellites do not sit motionless in orbit. Operators track close approaches constantly and perform avoidance maneuvers routinely. The point is not that collision is imminent under normal operations. The point is that the system now depends far more heavily on continual coordination, accurate tracking, functioning communications and the ability of many different spacecraft to move when needed. If any of those things fail broadly—if a solar storm scrambles atmospheric predictions and throws satellite trajectories into uncertainty, if a cyberattack disrupts communications, if a software fault cascades across multiple operators—the time available to recover has become very thin.
The reason is simple arithmetic. Low Earth orbit is physically vast, but the usable lanes are not. Satellites cluster in similar altitude bands and inclinations because those orbits are valuable for communications, Earth observation and other services. When more spacecraft occupy the same shells, the number of close approaches grows quickly. Megaconstellations changed the equation. SpaceX's Starlink is the most visible example, but Amazon's Kuiper and multiple Chinese broadband systems point toward an orbital economy where tens of thousands of active spacecraft may become routine. Each new satellite consumes a small part of a shared safety margin. Each one concentrates risk in high-demand altitude bands where collisions are more likely and debris lingers longer.
The debris problem is what makes the scenario genuinely consequential. Low Earth orbit is not populated only by working satellites. NASA's Orbital Debris Program Office estimates hundreds of thousands of marble-sized fragments and more than 100 million smaller objects already in orbit. A collision is not an isolated accident. The 2009 impact between the active Iridium 33 satellite and the defunct Russian Kosmos 2251 destroyed both spacecraft and created a debris population that other satellites have had to avoid ever since. Each serious fragmentation event produces objects too small to maneuver but large enough to damage or destroy something else. Some debris falls back quickly. Some remains aloft for years or decades, depending on altitude, size and solar activity. A dead satellite, a spent rocket stage or a small fragment cannot respond to a warning.
The paper does not claim that low Earth orbit is days away from a cascade. It does not say normal operations have failed or establish the precise moment at which a Kessler-type chain reaction would begin. The CRASH Clock depends on modeling choices, object catalogs and assumptions about collision severity. The authors present it as a tool for quantifying stress under a specified loss-of-control scenario, not as a perfect forecast. But the underlying pressure is visible and real: low Earth orbit now carries many more active spacecraft, many of them concentrated in high-demand altitude bands, and the recovery time from a broad system failure has shrunk dramatically.
What matters is not whether 5.5 days is exact to the hour. What matters is that the direction and scale of change are telling us something real about how the orbital environment is being used. Low Earth orbit is becoming infrastructure in its own right. Like airspace, shipping lanes or radio spectrum, it requires coordination to remain useful. The difficulty is that no single operator owns the whole environment. Every new spacecraft consumes a small part of a shared safety margin. Every collision avoidance system depends on data quality outside its own control. Every dead object left in a crowded shell becomes someone else's problem for as long as it remains there. The orbital region has not stopped working. But it now works because many moving parts keep working together, constantly, in a region where one bad collision can leave behind consequences that do not simply go away.
Notable Quotes
The CRASH Clock is not a schedule. It is a way of measuring orbital stress through time.— Thiele, Heiland, Boley and Lawler, authors of the CRASH Clock study
Low Earth orbit now works because many moving parts keep working together, constantly, in a region where one bad collision can leave behind consequences that do not simply go away.— The research paper's conclusion on orbital infrastructure dependency
The Hearth Conversation Another angle on the story
What does the CRASH Clock actually measure? Is it predicting a collision?
No. It's measuring how much time the system would have to recover if something went wrong broadly—if operators lost the ability to track objects or coordinate maneuvers. It's a stress test, not a prediction.
So 5.5 days means what, exactly?
It means if satellites couldn't dodge each other and operators couldn't see where things were going, the model suggests a damaging collision would become likely within that window. In 2018, the same scenario gave you 164 days. That's the change that matters.
Why did it shrink so fast?
Megaconstellations. Thousands of satellites launched into similar orbits in just a few years. The usable altitude bands are crowded now. More spacecraft in the same shells means more close approaches, less room for error.
But satellites are actively avoiding each other right now. So what's the real danger?
The danger is if the system that enables avoidance breaks. A solar storm could scramble atmospheric predictions. A cyberattack could disrupt communications. A software fault could cascade. In a sparse environment, you have time to fix it. In a dense environment, you don't.
What about all the debris already up there?
That's the trap. Every collision creates more debris. Debris can't maneuver, can't respond to warnings. It just stays there for years, creating hazards for everything else. One bad cascade could make entire altitude bands unusable.
Is this paper saying we've made a mistake launching all these satellites?
It's saying the scale of deployment is an environmental fact now, not just a business plan. The question isn't whether satellites are useful. It's whether the shared space that makes them useful is being treated as if its safety margin were unlimited.