For over thirty years now, most of the spy satellite photos the U.S. has obtained have been digital, broadcast to earth from space. This has gradually overloaded the ability of human analysts to thoroughly study all the data coming back. Since everything associated with spy satellites is highly classified, the problems with analyzing, or even just storing, the huge quantities of data, have been largely hidden from view. But there are warehouses full of tapes (the easiest way to store digital data), some of them decades old, and never seen by human eyes. In the last decade, there's been a growing flood of data coming from closer to the ground. UAV, droid and security cam video, airborne reconnaissance (both photo and electronic) and even troops mounting vidcams on their vehicles (to see what they might have missed during a patrol). It's been a growing crises.
In the last few years, more efforts have been made to make a lot of this data available on networks (not the Internet, but similar closed government intelligence nets), and to introduce software that can assist in the analysis. In the last decade, there have been spectacular advances in software that can examine digital data (photos, or equipment, like radar, transmissions) and find patterns. This is becoming the tool of choice to take a first look at most of the data collected. There is still a huge backup, but now there is at least a, theoretical, chance of catching up. And finding important things that, in the past, were simply passed over because no one had the time to look.
It all began when the first of these digital satellites, the KH (Key Hole) 9 was launched in 1971. These birds were large, nearly 15 tons, about four times the size of the earlier models that sent film back in canisters, to be developed. The KH-9s were nicknamed Big Bird. This one still sent back some film (of high resolution pictures) in canisters, but in 1977 the fifth generation dawned with the launch of the KH 11. This was a major breakthrough, as canisters were no longer needed and resolution was such that objects 70mm (a few inches) in size could be identified from 200 kilometers up via digital photos. The KH 11 telescopic cameras operated like a high resolution TV camera. Images were captured continuously and transmitted to earth stations. Computers were used to finish the process and produce photos identical to those taken by a conventional camera. You could even have motion pictures, as well as indications of heat and the nature of the various items. KH-11 could often tell what kind of metal an object on the ground was made of. This did not come cheap, these birds cost over $400 million each and lasted three or four years, depending on fuel usage. Moreover, you needed two of them up at the same time in order to guarantee coverage and save the birds from having to change orbit too frequently.
The next generation, the KH-12, was supposed to have been launched in 1987. Because of the space shuttle problems, only a belated KH-11 was launched in October, 1987. The KH 12 has several advantages over the KH 11. Along with improvements in ground data processing equipment, the KH 12 could send back data in real time. You could watch events on a large, high resolution screen as they were happening. This would also allow military headquarters and other users to get their satellite information directly, without going through a CIA or NRO (National Reconnaissance Office) processing center. Data from the more esoteric sensors would still have to be studied by the specialists elsewhere. The KH 12 was expected to make users even more enthusiastic about satellite reconnaissance. It did, but the flood of photographic and electronic data was still growing far larger than the force of analysts needed to make something of it. In addition to the KH series birds, there were radar and SIGINT (Signal Intelligence) satellites constantly broadcasting data. Then there are the Defense Support Program satellites, that use heat sensors to locate the hot plumes of missile launches.