Tomorrow Wars

Tomorrow Wars Volume 1 Issue 18: Space Case


40°27'00.9"N 80°14'03.6"W 900,00m

It is hard to hide a disaster in space.

The useful part of the heavens, the orbits that hold sensors pointed down at earth, are getting crowded. Bound by gravity and inertia, satellites orbit until they don’t, ideally plummeting towards earth, rocketing away into less-useful parts of space, or drifting dead in orbit until something changes.

In the six months I spent reporting and writing about the mathematical debate over satellite collision risk in orbit, it was hard to put in concrete terms what the event that ends orbit as we know it might look like.

And then, on Jan. 29, it almost happened.

I’m Kelsey D. Atherton, reporting from Albuquerque, New Mexico, and I am here this fortnight to talk about trash in space.

It is hard to talk about the dramatic stakes of a near miss, but the space press captured it well. Estimates of the collision risk fluctuated as more and more detail about the nature of the satellites pored in, making concrete a claim I first heard months prior.

Here is the line from the academic abstract that set this off: “A counterintuitive phenomenon has emerged in the conjunction analysis literature, namely, probability dilution, in which lower quality data paradoxically appear to reduce the risk of collision.”

Over months of back-and-forth in clarification, I was able to better parse the relationship between risk (which is fixed even if it is not known) and understanding of that risk (which hinges a lot on probability models and data quality). But with the close passage on the 29th, it was all cast into sharp relief: when observers knew only some details about the satellites, like their paths, the risk estimate was present but minimal. When it was revealed that one of the satellites, possibly a long-since decommissioned reconnaissance satellite from the 1960s, had booms extending almost 60 feet from its body, the collision estimates jumped drastically, taking risk from a 1-in-1000 chance to a 1-in-20. This is the uncertainty of orbit!

There is more, much more, to the risk of collision in space. Debris from one damaged satellite can in turn collide with other satellites. Tracking tools still struggle with some debris sizes, leaving many smaller bits of potential harm zipping around orbit unidentified and untracked. Anti-satellite weapon demonstrations blast debris up into useful orbits, and should any be used in anger, the debris fields that missiles will create from the torn and exploded husks of existing satellites will leave orbit a dangerous mess for years to come.

I hope, at least, to have found a way to make this danger somewhat intelligible, to have turned the stakes and risk of orbital flotsam into a clear portrait of what remains to be known.

Squaring this debate over risk, which has so far mostly taken place within academic journals, with the present optimistic plans for satellite launch in the 2020s is somewhat vexing. It is, in part, a tragedy of the commons, the kind of hard problem solved less by technological innovation and more by the begrudging work of meetings and compromise.

The scale of tragedy foreseen in academic debate is at odds with the present accelerating pace of launch. So far in 2020, SpaceX has already launched 180 new satellites, and the number of functional objects in low earth orbit could double by the end of the year.

Should a disaster come to orbit, it will be visible in the night sky. Amateurs and professionals, telescopes pointed upwards, will see the pieces that catch starlight, and can track the motions of new observable phenomena.

The “why” of the disaster will be harder to discern, at first. If space is a shared domain of peace, why was it not better maintained? If space is a home of invaluable military assets, how did it become so threatened by chance and trash?

Every near-miss is a possibility to breathe a sigh of relief that those questions do not, yet, require an answer. They will someday.


Cars do not at present transmit their locations, but a smartphone running a navigation app is a pretty good proxy, and aggregated that information can help drivers make useful choices about driving to avoid traffic . Until, that is, someone puts 99 smartphones running google maps into a little red wagon and runs it down the streets of Berlin. Simon Weckert’s digital art experiment is also a useful case study in the difference between what modern tools purport to offer, and what they actually offer users relying on them for navigation.


Barring treaties, the best way to stop a hypersonic arms race is to engineer weapons that can intercept other hypersonic problems. It is, to put it mildly, a hard problem, but one DARPA is investing in, as a way to build defenses against hypersonic cruise missiles.


To the extent that humans are good at commanding robots, it is because those techniques have been refined in video games. The genre of real time strategy, with players focused on commanding automatons to perform specialized functions, is mostly a diversion for the keyboard-and-mouse set, but it has also led to some innovation in controlling swarms of robots. By leaning on the tools developed for games, designers working on DARPA’s swarm project have made a more intuitive interface for sending teams of robots on scouting missions.


Today’s backronym is OFFSET, from DARPA’s “OFFensive Swarm-Enable Tactics” program. Why OFFSET instead of O-SET? I can only speculate, but it might have something to do with wanting to sound like the Third Offset Strategy. Or, perhaps, it is part of a concerted effort in sounding as unlike made-for-TV boy band O-Town, a reference that is now old enough to have enlisted and be on its second tour already.

That’s all for this week, my Tomorrow Warthogs. Any thoughts on the dangers that lurk in orbit, email me at

Recommended for you
Around The Web