This is what Russia’s Cosmos 2612 approaching to about 160 metres might have looked like from on board ICEYE-X36. It’s a simulation from public data, with the relative position of both satellites, their place above the Earth, the sunlight and the whole scene around them computed. We put the whole pass into our simulator, with orbital data from the last three weeks and our own cameras standing in for onboard ones. At the start the Russian satellite is on the far side of the Earth, several thousand kilometres away. When it comes out from behind the horizon it’s a pixel, maybe two, in every camera. With each manoeuvre it turns into a clearer shape, until it is a few hundred metres from @iceye_global and you can see every detail. What happened, briefly. From late September Cosmos 2612 manoeuvred into the orbit of ICEYE-X36, a radar satellite from the company whose imagery Ukraine uses. For many hours it circled ICEYE on a tight ellipse, from a few kilometres down to a few hundred metres. Overnight Saturday to Sunday it made a manoeuvre, raised its orbit and started drifting away. The 26 metres figure going around today is an automated prediction computed from pre-manoeuvre data. By our calculations the closest approach was probably about 160 metres, though the accuracy of public data allows anything from tens to hundreds of metres. Russia has done this kind of thing before, though in LEO usually at kilometres or tens of kilometres. A few hundred metres from a commercial satellite supporting one side of an ongoing war is something I struggle to find a public precedent for. There is one more thing about that circling. Cosmos went around ICEYE once per orbit, about every 95 minutes, while the Sun went around the pair of them. So in every loop it saw ICEYE from every side and in every lighting condition, from full light to side light, where shadows bring out the geometry of the antennas. Hard to design a better setup for optical reconnaissance. Quite possibly radio reconnaissance too. From a few hundred metres you can collect the characteristics of ICEYE’s signals that would be useful for future jamming, or even test it. According to the reconstruction, the departure manoeuvre lands almost exactly on the moment ICEYE came out of Earth’s shadow. I just don’t believe in coincidences like that. TBH: These reconstructions and our numbers are good to plus or minus some fifteen minutes, so we can’t be certain. The manoeuvre was reconstructed from public data by @AdamCohen8, an engineer in the @USSpaceForce, in his own private analysis. We get the same picture: right after the manoeuvre, around 01:38 UTC over Greenland, Cosmos was about 160 metres above ICEYE. The cameras in the video are the wide-angle @Scanway_SA SCS, from the family that filmed Ariane 6’s first flight, and SOP telescopes, from the smallest 3U/50, on orbit since 2023, through SOP120 from the family flying in the PIAST military constellation and on lunar missions, to SOP200 and SOP450. The biggest difference between them shows at a dozen kilometres or so, where SOP450 already resolves the satellite’s shape and the smallest camera still sees a few pixels. From a few hundred metres each of them sees detail. This is research-only analysis, based solely on public data and not coordinated with ICEYE. There are no Scanway cameras on ICEYE-X36. A few numbers from this simulation, for those who like that sort of thing: - for this scenario the simulator loaded about 380,000 stars from the part of the sky the cameras were pointing at; almost 58,000 fell within their fields of view, and 546 are bright enough to show up in the frame at the real exposure time - for everyday work it can draw on a catalogue of hundreds of millions of stars - the motion of both satellites comes from 159 sets of public orbital data spanning 24 days, with 6 clear Cosmos manoeuvres in the data - the video has 1,371 frames, with almost 4,000 renders behind them - the simulator holds over 35,000 objects in orbit, about 22,000 of them with known shape and size I think this is only the beginning. Commercial satellites work more and more for the military, which makes them targets. For now mostly for someone who flies up, takes a look and makes it clear they could do more. There will be more provocations like this in orbit, and operators, civilian ones too, will want to know what is happening next to their satellites and see it with their own eyes. Looking at it from the other side: if Polish and NATO satellites carried inspection cameras, every approach like this would become intelligence for us. Instead of guessing from orbital data what is flying next to us and what it is for, we would have close-up images of the visitor, and maybe traces of what it is trying to jam. At @Scanway_SA this is part of our day job. In early October we signed a contract with a European company for a series of 14 telescopes for on-orbit security applications, and a month ago, together with Revolv Space, we launched KIMBALL, an inspection camera on a 360° pointing mechanism with narrow and wide-angle optics in one unit. It can follow a target regardless of how the whole satellite is oriented, and on an approach from a few hundred metres, with the target sliding across the sky, that makes a huge difference. I have so many questions for @rmodrzewski about how ICEYE handled this operationally. I won’t ask them, since most of the answers are probably classified. What we don’t know: - positions from public data are accurate to hundreds of metres, and the manoeuvre itself is known only from reconstruction - ICEYE confirmed it moved its satellite; the data shows a small manoeuvre on the morning of 18 October, - there are no images of Cosmos 2612; the radar catalogue only tells us it’s a metre-class object, so its look in the video is a hypothesis based on what an inspection mission would need - its purpose is a hypothesis too, optical and radio reconnaissance alike The video is a simulation and is marked as one in every frame.