You think you have traffic problems? At least the cars around you have brake lights, marked lanes, speed limits and a generally understood set of rules about who has the right of way. Several hundred miles above us, tens of thousands of satellites and pieces of debris circle Earth at roughly 17,500 miles per hour. Many complete an orbit every 90 to 120 minutes. There are no fender benders at those speeds. A collision can turn two spacecraft into a cloud of high-velocity fragments, each capable of damaging something else.

A recent Wall Street Journal visualization put the scale in terms that are difficult to ignore: More than 34,000 satellites and pieces of debris are moving through low-Earth orbit, and more than 10,000 of those objects are related to SpaceX’s Starlink fleet. That count only includes objects large enough to catalog. Smaller fragments remain out there as well, moving too quickly to dismiss and often too small to track with enough precision to avoid.

ANIMATION 1 — THE TRAFFIC ALREADY THERE

None of this is an argument against satellites. They connect remote communities, guide ships and aircraft, monitor weather, support scientific research and underpin national security. Starlink has also shown what becomes possible when launch costs fall and a company can operate a constellation at unprecedented scale. The problem is that our capacity to place hardware in orbit has outrun our capacity to govern the traffic that hardware creates. What we see today is the relatively quiet version of what is coming.

The Next Wave is Already Forming

The Journal, citing Dallas Kasaboski of Analysys Mason, reported that between 2026 and 2034 at least 41,000 more satellites are expected to be deployed to low-Earth orbit. About two-thirds are expected to come from SpaceX, Amazon and Chinese companies. SpaceX continues expanding Starlink. Amazon Leo, formerly Project Kuiper, is building a constellation of more than 3,000 satellites and has launch agreements with Arianespace, Blue Origin, SpaceX and United Launch Alliance. China is developing Guowang, Qianfan and other networks, with plans that extend into the tens of thousands of satellites. India is evaluating sovereign low-Earth-orbit networks of its own. The space economy is no longer a contest among two superpowers and a handful of government agencies. It is becoming a crowded combination of national programs, commercial operators and strategic infrastructure.

Then come orbital data centers. In January, SpaceX filed an application with the Federal Communications Commission for a system of up to 1 million satellites. Starcloud requested authority for as many as 88,000. Blue Origin followed with Project Sunrise, a proposed constellation of up to 51,600 satellites. Those three filings alone describe maximum systems totaling 1,139,600 satellites.

That number requires an enormous asterisk. A regulatory filing is not a launch manifest, and a proposed ceiling is not a forecast. Many of these satellites may never be financed, built or launched. The engineering and economics of operating AI infrastructure in orbit remain daunting. Still, the filings show the scale at which serious companies are beginning to think. Even a small fraction of those plans would be added to an environment whose coordination mechanisms are already being tested.

ANIMATION 2 — THE LOOK AHEAD

We Solved Launch. We Did Not Solve Traffic

Orbit is not a parking lot in which every satellite sits at the same altitude. Spacecraft occupy different orbital shells and inclinations, and operators intentionally separate them. But their paths cross. Predictions carry uncertainty. Dead satellites and debris cannot maneuver. An avoidance move by one active spacecraft changes its future conjunctions with other objects. As the population rises, the number of calculations, warnings and decisions rises with it.

There is no universal orbital right-of-way rule. Operators are not all required to share equally precise location data or maneuver intentions in the same format. Different tracking networks and commercial services can produce different estimates of the same encounter. Military spacecraft introduce legitimate secrecy concerns. Coordination has improved, but too much still depends upon voluntary cooperation among operators using systems built by different governments and companies. We are building the interstate highway system of space without agreeing which side of the road anyone should drive on.

We already know what a collision looks like. In 2009, the functioning Iridium 33 communications satellite and the defunct Russian Cosmos 2251 satellite struck one another. NASA researchers reported that the collision produced more than 1,800 debris objects approximately 10 centimeters or larger. That was a serious accident, but it did not produce a comprehensive global traffic regime. The uncomfortable lesson is that an ordinary satellite collision may no longer be enough to force political action.

A Global Commons Governed by a Patchwork

It would be wrong to say nobody is working on the problem. The U.S. Commerce Department is developing the Traffic Coordination System for Space, known as TraCSS, to provide civil and private operators with space-situational-awareness information. As of August, it reported 70 pilot users representing more than 11,345 satellites, along with national government accounts for 10 countries. Europe operates EU Space Surveillance and Tracking. Militaries, national agencies and commercial companies maintain additional tracking and warning capabilities.

The United Nations adopted 21 guidelines for the long-term sustainability of space activities in 2019. They cover regulation, operational safety, international cooperation and scientific research, but they are voluntary and not legally binding. ISO 23705:2026, published in June, establishes workflows, data requirements and technical guidance for identifying, evaluating and avoiding orbital collisions. That is valuable technical plumbing. A standard, however, is not the same thing as an enforceable traffic law.

The 1967 Outer Space Treaty makes countries internationally responsible for their national space activities, including those conducted by private companies, and requires authorization and continuing supervision. It also establishes liability principles for damage caused by space objects. What it does not provide is a real-time operating system for deciding which of two spacecraft should maneuver, when it should move or what information must be exchanged first.

The result is a collection of important pieces that does not yet add up to a global system. The solution cannot be SpaceX’s system, America’s system, China’s system or Europe’s system. It must be a system in which all of those systems can communicate, predict, coordinate and, when necessary, yield.

Global Does Not Mean One Control Room

A global solution does not require handing control of every satellite to an enormous bureaucracy in New York or Geneva. In fact, the most practical model is already beginning to emerge. TraCSS has described a federated vision consisting of national and regional space-awareness hubs. The U.S. and EU systems are studying how to share information and align services, while recognizing that global safety requires more providers to join.

That federation needs a mandatory minimum operating layer. Every participating system should contribute to a continuously updated catalog, use common data formats and exchange accurate ephemerides and maneuver intentions. Operators need an agreed protocol for who moves when two active spacecraft are on a collision course, with sensible priority for crewed vehicles and objects that cannot maneuver. Sensitive military assets will require protected channels that communicate enough safety information without exposing every mission detail. Serious near misses should be reported and independently investigated so that the entire system learns from them.

The same compact should establish enforceable end-of-life and deorbit obligations. Operators should carry insurance or provide financial guarantees sufficient to cover responsible disposal and foreseeable damage. Compliance does not need to depend on imaginary space police. Governments can attach it to launch and operating licenses. Regulators can connect it to spectrum access. Insurers, lenders and government buyers can make participation a condition of coverage, financing and procurement. Countries can require foreign operators seeking access to their markets to meet the same baseline rules.

Any credible agreement must include the United States, China, Europe, India and Russia, as well as the largest commercial operators. Leaving out a major space power would reproduce the fragmentation the compact is supposed to cure. The UN Committee on the Peaceful Uses of Outer Space can supply legitimacy and a forum. ISO can provide a common technical grammar. TraCSS, EU SST and comparable systems can become interoperable hubs. National governments can supply enforcement through the authorities they already possess.

None of that requires Washington, Beijing, Brussels, New Delhi and Moscow to settle every argument they have on Earth. Rivals do not need to trust one another completely to agree on the mathematics of avoiding a collision. They need verifiable data, predictable procedures and confidence that everyone will follow the same rule when seconds matter.

Must Catastrophe Come First?

History is not reassuring. The Titanic sank in 1912; the first international Safety of Life at Sea convention was adopted two years later in response. In 1956, TWA Flight 2 and United Airlines Flight 718 collided over the Grand Canyon, killing all 128 people aboard. The Federal Aviation Administration calls that disaster a catalyst for major changes to air-traffic control and the creation of the Federal Aviation Agency in 1958.

We have a habit of writing safety systems in wreckage and naming reforms after the people who died before politicians acted. Do we really need an orbital Titanic before the world writes the traffic rules for space?

Perhaps another collision between two uncrewed satellites would still not be enough. The catalytic event might be a debris cloud that disables several spacecraft, the prolonged loss of communications or weather services, or an encounter that threatens a crewed vehicle. Worse, an unexplained maneuver or sudden satellite loss could be interpreted as hostile action before investigators establish that it was an accident. Space traffic management is therefore more than an environmental or commercial concern. It is also a means of preventing misunderstanding and escalation among nuclear-armed states.

ANIMATION 3 — AVOID THE COLLISION

The real global space race is not simply about which company or country can put the most hardware into orbit. It is whether we can build the rules quickly enough to preserve the shared environment on which every one of those satellites depends. Commercial ingenuity can build better sensors, prediction systems and autonomous maneuvering technology. It cannot, by itself, decide what every operator and every country is required to do.

We should not have to watch two spacecraft collide at 17,500 miles per hour, or lose a crew, to learn that orbit needs traffic rules. We already know what must be built. The question is whether we build it before the wreckage forces us to.