Somewhere in low Earth orbit, a Starlink satellite fires its krypton thrusters and nudges itself a few hundred meters sideways. Less than two minutes later, another one does the same thing. Then another. According to figures SpaceX reported in its semi-annual orbital debris filings with the Federal Communications Commission, the company’s constellation executed roughly 144,000 collision-avoidance maneuvers over a recent six-month stretch, a rate that works out to about one dodge every 1.8 minutes, day and night, without pause.
The number, disclosed as part of SpaceX’s regulatory obligations under the FCC’s updated orbital-debris mitigation rules, captures something that no single satellite operator has ever had to manage before: a fleet so large that avoiding collisions has become an industrial-scale, around-the-clock operation. As of early 2026, SpaceX operates more than 6,700 active Starlink satellites, making it by far the largest constellation in history. And the company has FCC authorization to eventually fly more than 12,000.
How the warning system works
Every maneuver begins with a detection. The U.S. Space Force’s 18th Space Defense Squadron at Vandenberg Space Force Base runs conjunction screenings multiple times per day, scanning the cataloged population of satellites, rocket bodies, and debris fragments for objects whose orbits bring them dangerously close to one another. Those screenings feed into NASA’s Conjunction Assessment Risk Analysis (CARA) program, which calculates the probability and potential severity of each flagged encounter and pushes warnings to satellite operators.
Detection is largely automated. The decision to act is not. When a CARA alert lands, the operator must weigh the collision probability against the cost of maneuvering: propellant burned, mission time lost, and the downstream effect on neighboring satellites whose orbits may shift in response. For a single science satellite, that calculus happens a few times a year. For Starlink, it happens hundreds of times a day.
What the number does and doesn’t tell us
The 144,000 figure is striking, but it comes with important caveats. SpaceX has not published a granular breakdown of what triggered each maneuver. A small altitude tweak to widen the miss distance with a tracked debris fragment and an emergency burn to avoid a defunct rocket stage both count as maneuvers, yet they represent vastly different threat levels.
Per-satellite averages also obscure the real distribution. If 6,700 satellites shared the total evenly, each one maneuvered roughly 21 times over six months, or about once every eight or nine days. But orbital mechanics makes an even split unlikely. Satellites in the most congested shells, particularly the 550-kilometer band that hosts the bulk of the first-generation fleet, almost certainly maneuver far more often than those in less crowded altitudes. Proximity to known debris clusters, orbital inclination, and the density of other operators’ hardware all influence how frequently a given spacecraft triggers an alert.
Perhaps the biggest gap is the absence of any public breakdown separating external threats from so-called self-conjunctions, close approaches between two Starlink satellites within the same constellation. As the fleet grows, intra-constellation geometry becomes an increasingly significant source of alerts. Without that split, it is impossible to know how much of the maneuver burden is driven by outside traffic and how much is a consequence of SpaceX’s own constellation design.
Why the math gets worse, not better
Orbital conjunction risk does not scale linearly with the number of objects in a given altitude band. Because screening algorithms must evaluate every possible pairing, the number of potential close approaches grows roughly with the square of the population. Double the satellites in a shell, and you roughly quadruple the conjunction pairs that need checking. That combinatorial reality is why even modest constellation expansions can produce outsized jumps in maneuver counts.
SpaceX is not the only operator adding hardware. Amazon’s Project Kuiper began launching its first production satellites in 2025, targeting an eventual constellation of 3,236 spacecraft in overlapping altitude bands. OneWeb, now part of Eutelsat, operates more than 600 satellites. China’s Qianfan (G60) constellation has begun deployment with plans for thousands more. Each new entrant thickens the traffic in low Earth orbit and multiplies the screening workload for every other operator already there.
The European Space Agency’s most recent Space Environment Report documented a sharp year-over-year rise in conjunction alerts across all operators, a trend ESA analysts attributed primarily to constellation growth. The report noted that the number of tracked objects in low Earth orbit has more than doubled since 2020, driven largely by mega-constellation deployments.
The burden on everyone else
For smaller operators, the consequences are concrete and immediate. Every conjunction alert demands staff time, computational resources, and, if a maneuver is ordered, propellant that cannot be recovered. A university-built CubeSat with a few hundred grams of fuel cannot absorb dozens of avoidance burns without sacrificing its science mission. National space agencies with larger budgets still must now bake conjunction avoidance into mission planning from day one, a cost that barely existed a decade ago.
NASA uses the CARA system to protect the International Space Station, the Hubble Space Telescope, and its fleet of Earth-observation and science satellites. The agency has publicly noted that the volume of conjunction warnings it processes has grown substantially, requiring more analysts and more computing power to maintain the same level of safety. When the ISS performs a debris-avoidance maneuver, the crew sometimes shelters in their return vehicles as a precaution, a reminder that the stakes in orbit are not abstract.
Governance that hasn’t kept pace
Today, conjunction warnings flow through a patchwork of military tracking, civil risk analysis, and voluntary operator-to-operator coordination. There is no single global authority that can compel a maneuver or resolve conflicting plans when two spacecraft both want to occupy the same safe corridor. The FCC’s 2022 decision to shorten the post-mission deorbit window from 25 years to five was a step toward reducing long-term debris accumulation, but it does not address the real-time traffic management problem.
International guidelines on debris mitigation and end-of-life disposal, maintained by the Inter-Agency Space Debris Coordination Committee and endorsed by the United Nations, remain largely voluntary. Enforcement mechanisms are weak, and compliance is uneven. In that regulatory vacuum, large operators’ self-reported metrics on collision avoidance carry outsized weight, because they are among the few quantitative indicators of how congested key orbital bands have actually become.
The U.S. Department of Commerce has been working to stand up a civil space traffic coordination capability, a role currently filled by the military, but progress has been slow and funding uncertain. Until a civilian system is operational, the Space Force remains the primary source of conjunction data for American and allied operators, a dual-use arrangement that some analysts argue creates conflicts between national security priorities and transparent traffic management.
What 144,000 maneuvers actually signal
There are two ways to read SpaceX’s number, and both are probably true at the same time. Frequent avoidance burns are evidence that the system is working: threats are detected, warnings are issued, and satellites move out of harm’s way before anything collides. SpaceX has invested heavily in autonomous maneuvering software, and the company’s ability to execute thousands of burns per month without a collision is, by any measure, an engineering achievement.
But a sustained pattern of maneuvers at this tempo also signals that the underlying traffic density in low Earth orbit is approaching a level where routine operations become increasingly fragile. Each burn consumes krypton propellant that shortens a satellite’s useful life. Each maneuver shifts one spacecraft’s orbit in ways that can cascade into new conjunction alerts for neighbors. And each decision not to maneuver carries a residual collision risk that, multiplied across tens of thousands of objects, compounds over time.
Until more detailed, independently verifiable data is released, the 144,000 figure should be treated as an important but incomplete snapshot. It tells us the scale of the problem without fully describing its shape. What it makes unmistakably clear is that low Earth orbit is no longer an empty frontier. It is a busy, shared environment where the rules of the road are still being written, and the traffic is arriving faster than the rulemakers can keep up.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
