Three Days to a Collision: Why Space Junk Is Nearing a Breaking Point

"We might not see a substantial change during our own generation, but we will see a steady increase in risks and collisions," says Hugh Lewis, an astronautics professor at the University of Birmingham.

Starlink satellites use ion thrusters and autonomous AI guidance to dodge close passes. The industry standard triggers a maneuver at a 1-in-10,000 collision chance; SpaceX now acts at just 1 in 3.3 million, performing more than 400,000 avoidance maneuvers per year, according to figures the company reported in July 2026. In January, SpaceX also announced it would shift most satellites 13% below their original 340-mile (550 km) altitude, where fewer objects exist and failed satellites deorbit faster.

OneWeb operates higher, at 750 miles (1,200 km), where drag is weaker. A failed object there could stay aloft for centuries. Both companies deorbit satellites with their own propulsion after roughly five-year lifetimes — but if propulsion fails, the satellites become junk.

What is changing the odds?

Two external factors now pull orbital lifetimes in opposite directions. Greenhouse gases warm the dense troposphere but cool and thin the mesosphere above it, reducing drag at altitude. Lewis's team found this could quintuple the lifetime of some debris by 2100.

Solar cycles work the other way. "In the critical range of 300 to 500 kilometers, the density of the atmosphere is very dependent on whether we're at solar maximum or solar minimum," says astrophysicist Jonathan McDowell, who has catalogued launches and reentries for decades. Each 11-year solar maximum heats and expands the atmosphere, increasing drag. The Hubble Space Telescope loses altitude at every solar maximum, each loss bigger than the last; without intervention, it could come down in the next one.

Where is the risk concentrated?

SpaceX has asked the Federal Communications Commission to approve space-based data centers with 1 million satellites orbiting above Starlink — requiring roughly 200,000 launches a year. At SpaceX's reported 1% failure rate, "you're replacing 2,000 satellites a year," Lewis says. "Those failed satellites alone would meet the critical Kessler threshold, without even looking at the million satellites." In January, a Russian spy satellite fragmented unexpectedly in a high-orbit graveyard region, possibly from an impact.

Darren McKnight, senior technical officer at LeoLabs — which coordinates about a dozen ground-based tracking radars — points to the danger zone around 975 km (600 miles), where 1970s- and 80s-era rocket bodies and defunct payloads sit without any avoidance capability. "If we have a heart attack, it's going to happen there," he says. Removing the 10 worst objects at 840 and 975 km — three defunct satellites and seven rocket stages — would cut debris-generating potential by 30%, offering what he calls "much better return on investment" in mitigation.

Is anyone cleaning up?

The only completed cleanup to date came in 2022, when China's Shijian-21 towed a defunct Beidou-2 G2 satellite from geostationary orbit to a graveyard orbit 1,865 miles (3,000 km) higher. In 2025, Shijian-25 docked with fuel-depleted Shijian-21, likely testing refueling — a maneuver U.S. Space Force officials labeled a "priority watch" event. Japan's Astroscale has flown two LEO test missions and next aims to retrieve a retiring OneWeb satellite; Switzerland's ClearSpace partners with the European Space Agency; California-based TransAstra, which tested a capture bag aboard the ISS, hopes to launch in 2028 to snag a 100-ton asteroid with technology adaptable to debris.

"I am a fan of space exploration," McDowell says, "but I want us to do it sustainably, in a way that doesn't trash the environment." With the space industry now worth more than $600 billion, some see profit in clearing the debris trail. The capacity that cluttered orbit is the same capacity that can clean it.