How the US Built Its Super-Secret Spy Satellite Program


Andrew Tarantola

Ethics aside, espionage is an indispensable part of statecraft. The ISR [Intelligence, Surveillance, and Reconnaissance] information gathered helps steer national policy decisions for everything from mundane trade negotiations to the blackest of ops. And nowhere is this more evident than in the development of the US spy satellite fleet during the Cold War. These orbital telescopes granted an unprecedented peek over the Iron Curtain—revealing Soviet military capabilities, supply reserves, industrial sites, and more—that no ground-based spook could hope to provide.

During the Cold War, accurately ascertaining the USSR's military capabilities was a top US priority—as well it should have been given that we had as many as 21,000 nuclear warheads pointed at each other during that time. And while we had plenty of spies operating in Moscow, the view from overhead provided the President and his cabinet key insights into the extent of Soviet strategic capabilities which influenced defense planning and arms control negotiations. As such, the US invested vast sums of money into high-altitude research—from early "weather balloons" to the SR-71 Blackbird and U2 Dragon Lady to orbital telescopes—and established not one but three Federal agencies—the National Reconnaissance Office (NRO), the National Security Agency (NSA), and the Central Intelligence Agency (CIA)—all in an effort to glean any speck of information that could give us an advantage.

Satellite technology is, by far, the most expensive ISR method at the US's disposal but also the most effective, its results well worth the billions of dollars spent. As President Lyndon B. Johnson famously quipped in 1967 after a Soviet hoax led to worries of a bomber gap:

I wouldn't want to be quoted on this ... We've spent $35 or $40 billion on the space program. And if nothing else had come out of it except the knowledge that we gained from space photography, it would be worth ten times what the whole program has cost. Because tonight we know how many missiles the enemy has and, it turned out, our guesses were way off. We were doing things we didn't need to do. We were building things we didn't need to build. We were harboring fears we didn't need to harbor.

Of course, much of the development of our national reconnaissance capabilities is still shrouded in veils of classification. Heck, the NRO was established in 1961 and operated for three decades before the government even ever acknowledged its existence. Press reports made limited references to the agency as far back as 1971, but it wasn't until the Deputy Secretary of Defense revealed the NRO in 1992, was it ever formally discussed by the DoD. Oversight from the DoD and Congress was virtually non-existent save for the "open-checkbook" policy of the times. As long as the intelligence justified the price tag, any cost was acceptable. It wasn't until the early 1990's that any information on these devices was declassified, after the fall of the Soviet Union brought an end to the Cold War. Even now, information on the early satellites is sparse and anything after 1972 is non-existent save for a few photos taken by the KH-11 satellite which were leaked to Jane's Defence Weekly in 1985.

What we do know is that the US has been researching high-altitude reconnaissance technology since about 1946 when the RAND project, precursor to Rand Corp., began campaigning for its development. When the Army and Navy couldn't agree on who would have control over the orbital technology, it was assigned to the newly-formed USAF in 1947. It took a few years for RAND researchers working on "Project Feedback" to figure out how a satellite would even function—this was a brand new technological concept, mind you—but by 1953 they had not only devised the general characteristics and capabilities of a reconnaissance satellite but had begun to develop many of the components as well, like the television system and altimeter. The Atomic Energy Commission also began work on miniaturized nuclear power sources for the vehicles at that time. By 1954, the USAF accepted RAND's assertion that the technology was of "vital strategic interest to the United States" and officially established the US satellite program.

The Corona Program

The first such program was the Corona project, a codeword itself code named "Discoverer" for the public explanation of why the government was firing a rocket into space (a rare event in the late 1950s that would have attracted a curious public and international scrutiny). The program began in 1959 at the Onizuka Air Force Station, ran until 1972, and was declassified in 1995 by President Clinton. Its initial budget was a modest $108.2 million ($860 million adjusted to 2013), though that quickly increased following the 1960 incident in which Gary Powers' U2 was shot down over Soviet airspace. The 144-member family of Corona satellites—each designated Keyhole-#, or KH-#, depending on the spacecraft iteration—were produced and operated by the CIA in conjunction with the USAF and provided invaluable photographic surveillance of the Soviet Union as well as the People's Republic of China, and other Communist countries.

Launched aboard a Thor booster rocket and Agena spacecraft, these satellites relied on a pair of five foot long stereoscopic Itek cameras using 12-inch, f/5 triplet lenses and a 24-inch focal length (later models also incorporated a third "index" camera for reference). The early cameras could achieve a 40-foot resolution. By KH-3, optical improvements decreased that figure to 20 feet. Later missions continued to halve the resolution until researchers were able to resolve one-foot wide objects, realized that that was way too close to be of any strategic use and backed off to a more manageable 3-foot resolution.

They were fed a special Eastman Kodak 70 millimeter film that produced 170 lines per mm—more than three times the 50 lines/mm resolution earlier WWII aerial photography could compose. The first Coronas carried a paltry 8,000 feet of film—per camera—though through improvements in the film chemistry and design reduced the material thickness, researchers were eventually able to double that amount. The cameras themselves underwent numerous upgrades as well, elongating to nine feet and incorporating panoramic Petzval f/3.5 lenses.

Once the camera had run through its full complement of film, it would eject the roll via a reentry capsule designed by General Electric. After the capsule discarded its heat shield at 60,000 feet, it deployed a parachute and could either be nabbed by a passing plane equipped with a claw hook (above) or land safely in the ocean where it would float for two days awaiting pickup. If the capsule wasn't retrieved within 48 hours, a salt plug at the bottom of the canister would dissolve and sink it. If it was picked up in time, the film would be transported to Rochester, New York, for processing at Eastman Kodak's Hawkeye facility.

The Argon Program

The KH-5 ARGON ran in conjunction with Corona from 1961 to 1964, though never with the same degree of success. These 1150 - 1500 kg satellites manufactured by Lockheed Martin and operated by the NR used a single 76 mm focal length camera with a 140 meter resolution were operated primarily for map-making—they were the first to image Antarctica from space—and took less than a week to produce. Of the 12 flights attempted, however, only five successfully put the unit in orbit.

The Lanyard Program

The KH-6 Lanyard program was the NRO's first attempt at high definition photography but lasted just six months and three launches in 1963, two of which failed to produce images. These 1500 kg Lockheed satellites were hastily constructed using the previously-cancelled Itek "E-5" camera in order to survey a rumored anti-ballistic missile site near Tallinn, Estonia. The E-5 had a 66-inch focal length and six foot resolution covering a 9 x 46 mile area. The only successful flight returned 910 photographic frames. However, the image quality was so poor that they were virtually useless.

The Gambit Program

Outside of the Corona program, America's initial attempts at satellite photo-reconnaissance failed more often than not. The KH-7 and KH-8 series, codenamed Gambit, were a marked departure from that trend and the only other predominantly successful satellite ISR program in the 1960s. This 3,000 kg Low Altitude Surveillance Platform developed by Lockheed flew just 75 miles up (Coronas orbited at 100 miles) and operated for nearly two decades from 1964 to 1984. No fewer than 54 such satellites launched (these things only worked for three months, tops) from Vandenberg AFB aboard Titan III rockets during that time.

Eastman Kodak's A&O Division in Rochester, New York, produced the Gambit's primary strip camera system. With a focal length of 175.6 inches, a 6.3 km wide coverage area, and 3-foot resolution, the KH-8 was ideal for gathering high-resolution images of Soviet sites. Unlike conventional aperture cameras, the Gambit's slit camera reflected light off of a 48-inch mirror, through a slit aperture, and on to a moving length of Eastman Kodak Type 3404 film. It would then either drop the roll as the Coronas did or automatically develop the photographs, scan them, and transmit the images back to Earth in as little as 20 minutes through the Film Read-Out GAMBIT (FROG) feature (though after $2 billion dollars and nearly a decade of development the 1971 administration nix(on)ed it).

In addition to keeping tabs on Soviet air capabilities, Gambit was also designed to photograph the spacecraft around it. This ability came in handy in 1973. The brand new Skylab had just launched when its meteoroid shield broke loose and damaged the space station. As NASA scrambled to send up a manned repair mission, the NRO launched a new Gambit, which snapped this picture and helped NASA engineers plan accordingly.

The HEXAGON Program

The KH-9 HEXAGON was, by all accounts, an unmitigated success with 19 of its 20 launches reaching orbit between 1971 and 1986. This $3.262 billion Lockheed-built NRO program is officially deemed a Broad Coverage Photo Reconnaissance satellite but is better known as "Big Bird." And while its existence wasn't revealed until 2011, the program dates back to the 1960s as a successor to the Corona program.

The first generation of HEXAGON employed a pair of f/3.0 folded Wright Camera cameras with a 60-inch focal length able to resolve objects down to 2 feet and carried four re-entry vehicles. The last three generations featured a pair of panoramic cameras as well as upgraded electronics, C&C systems and nitrogen-supplied re-entry canisters. They also a began surviving longer. Most spy satellites have very limited life spans—two to three months—and once they're out of film they have no further purpose. But with ever increasing film payloads, the final iteration of the KH-9 lasted 275 days in space. Between 1973 and 1980, these satellites imaged every square foot of the Earth in 29,000 pictures, much of it better quality than LANDSAT, a rival satellite mapping program. Most of these images have been declassified since 2002, though sensitive areas such as government installations and most of Israel remain tightly guarded.

The KENNAN Program

The KH-11 KENNAN is the most advanced recon satellite to be unclassified. First launched in 1976 by the NRO, it's the first US satellite to employ an EO digital sensor and charge-coupled device (CCD), which reportedly provides an Enemy of the State-style real-time observation capability. Very little is known about the satellite's hardware though many have speculated that its roughly the same size as the Hubble Space Telescope with a similar 2.4-meter mirror producing a six inch resolution. There's also wide speculation that the KH-11 is the source of images declassified in the wake of the 1998 embassy bombings, as well as others of China and Russia declassified the year prior. The images the CIA used to find Osama bin Laden's hideout were reportedly supplied by the KENNAN. Fifteen KH-11's have been launched in total—nine between 1976 and 1990 aboard Titan-3D rockets, five between 1992 and 2005 aboard Titan IVs, and the final one in 2011 aboard a Delta IV—at an estimated cost of $2.2 to 3 billion.

The end of the Cold War certainly put a damper on reconnaissance satellite funding, as did the rise of commercial satellite technology, but it remains a staple of our intelligence gathering resources. The technology has also found new use in providing tactical information to ground troops (not having to catch film canisters with sky hooks helps). Satellite imagery was first used in 1991 during Desert Shield and again in Iraq, Bosnia, Kosovo, and Afghanistan.

Be Proud of Hexagon

This article was written by Dwayne Day "Space policy analyst and historian who works in Washington, DC." It was written for the SpaceReview online publication.

Those of you who worked on Hexagon have every reason to be extremely proud of what you did. It was so important to the security of the United  States and how it helped keep the peace during the cold war. 


Black ops and the shuttle (part 3-2): The HEXAGON ghost haunting the desert storm

by Dwayne A. Day

Monday, December 18, 2017

In summer 1990, Saddam Hussein’s military forces invaded Iraq’s southern neighbor, the tiny country of Kuwait. The United States mobilized for war and in early 1991 launched a massive air attack against Iraqi forces, later followed by an invasion of both Kuwait and Iraq. Generally referred to as the Persian Gulf War, the two primary phases of the operation were designated Operations Desert Shield and Desert Storm. In relatively short time, the Iraqi military was defeated.

“Wide area synoptic coverage could have captured the whole Iraqi land mass and forces in one fell swoop and had the film and some exploitation to them in two days,” a former NRO official explained.

United States intelligence capabilities, including satellite assets, were used during the war to an unprecedented extent for a military operation, and Desert Storm earned the moniker “the first space war.” After the end of the conflict, several military commanders complained about the availability of satellite imagery. The House Armed Services Committee (HASC) produced a report in 1993 titled “Intelligence Successes and Failures in Operations Desert Shield/Storm.” In the report, one unnamed military commander stated:

There is a need for wide-area synoptic coverage. The area occupied by Iraqi forces was on the order of 27,000 to 30,000 square miles, the size of four New England states. In hindsight, getting rid of both the SR-71 (high-altitude photographic reconnaissance aircraft) and [a wide-area satellite imagery system] at the same time was short-sighted. The Commander-in-Chief [CINC] lacked synoptic coverage.

The report further stated “The absence of wide-area coverage has been compared to ‘searching New York City by looking through a soda straw.’”

The SR-71 was retired in spring 1990, only a few months before the Iraqi invasion of Kuwait. The last successful HEXAGON mission was in 1984, followed by the last launch in 1986. The unnamed military commander quoted in the HASC report was clearly bemoaning the loss of the incredibly powerful HEXAGON system, and unhappy with its replacement.

A former NRO official agreed with the commander’s assessment. “Wide area synoptic coverage could have captured the whole Iraqi land mass and forces in one fell swoop and had the film and some exploitation to them in two days,” he explained. He also thought that the soda straw analogy was misleading, because it missed the fact that the individual images stitched together to produce a mosaic of a larger area were taken at different times, and things could change in the interim. “How many times would you count the same taxi (tank) that had moved from place to place and been imaged multiple times at multiple places?” he asked. There really was no substitute for the kind of coverage that HEXAGON had provided, photographing vast amounts of territory very rapidly, seeing entire armies in a single pass overhead.

In his 1996 interview, NRO director Martin Faga had much to say about the NRO’s support to Operations Desert Shield and Desert Storm. Faga stated that he felt that much of the criticism of satellite capabilities during Desert Storm was unfair. “Over the period of the war, from the invasion in August through the end of the conflict in February, the number of satellites available to watch the area was the same, but the product almost doubled,” he said. Faga attributed this to NRO assistance in targeting and effective resource management.

Although Faga did not directly address the issue of broad area coverage raised by the military commander, he attributed most of the complaints about lack of imagery to a ground distribution problem, not a supply problem. The United States Army, Faga stated, had spent a significant amount of money on ground systems for processing satellite imagery and enabling its distribution to field commanders. In contrast, the Air Force had spent very little money on that capability. As the military build-up in theater occurred, General Norman Schwartzkopf, who commanded all allied forces, partially rectified the situation by sharing the Army’s capability with the Air Force, spreading the assets thin and leading to much grumbling. The problem was not the satellites, Faga argued, and he noted that afterwards the Air Force began investing significantly in ground processing and distribution equipment.

Although the distribution of intelligence information in the field had resulted in much criticism, notably, the quote in the House Armed Services Committee report identified collection systems—specifically the lack of both the HEXAGON and the SR-71—as a major problem. How the collection systems fit into the overall intelligence collection mission during the late Cold War was an even more complicated story.

Tracking tanks in the Gulf War was challenging given the limited imagery available to forces in the field. (credit: US Defense Department)

Order of battle

Several former senior CIA analysts were asked about their opinions concerning the end of the HEXAGON era and the deficiencies of the KENNEN for the search mission. “It was true,” one former analyst confirmed, the HEXAGON had been superior to the system that replaced it even though the HEXAGON used film and returned its images days—sometimes weeks—after they had been taken.

Imagery provided proof of what existed, but not what could exist in the next few years. And sometimes the imagery provided data that could be accurate but misleading.

Another analyst who was also familiar with the NRO disputed that assessment: “Virtually all relevant shortfalls in area coverage between the two could be managed with effective collection tasking management, which was problematic, given competing priorities.” But a former NRO official who was intimately familiar with the HEXAGON’s capabilities thought that the problem of competing imagery collection priorities was not an excuse, it was the whole point: without a dedicated satellite like HEXAGON providing search imagery, of course the search mission would be neglected for other missions that were considered higher priority by the intelligence community, and it was.

But the former analysts also added that imagery collection and its shortfalls was only part of a much bigger puzzle that they were trying to assemble. Imagery alone could not answer all of the questions that intelligence analysts were trying to answer. Imagery provided proof of what existed, but not what could exist in the next few years. And sometimes the imagery provided data that could be accurate but misleading.

For instance, during the 1970s and ’80s CIA analysts had to try and project Soviet tank production in coming years while also calculating what tanks were already deployed and ready to face down NATO in Europe. This created several complicated dilemmas, because analysts realized that the Soviets should have been producing far more tanks than the satellites were showing deployed in the Soviet Union and the various Warsaw Pact nations. The satellites had also photographed Soviet tank production factories, which were huge and should have been capable of turning out many tanks per year. Were the satellites not photographing all the extra tanks in the field—perhaps because the Soviets were very effectively hiding them—or was Soviet production much lower than the CIA’s analytical tools indicated?

Although it has been gone for decades, the HEXAGON’s capabilities have never been replicated.

It turned out that Soviet tank production was much lower than initially calculated. While the Soviets had vast factories dedicated to armored vehicle production, they were incredibly inefficient and produced far fewer tanks than American industry did based upon floor space and workforce. It was not just the size of a factory or the number of tanks in the field revealed by the satellite images, it was the analytical models used to interpret all that data that were required to produce more accurate intelligence that could be delivered to decision makers.

Later, while trying to verify the Treaty on Conventional Armed Forces in Europe in the early 1990s, American inspectors got into heated arguments with Russian field commanders who they thought were showing them obsolete vehicles rather what was officially in the Russian reporting and was supposedly in the inventory. The inspectors always heard the same response: “We are showing you what we have, not what we report to Moscow.” The Soviet system was not only one big puzzle to the Americans, it was a puzzle to the Soviets themselves.

Although the satellites could not answer all the questions, these incidents also emphasized the value of satellites over spies. Even if a spy got top secret information from inside the Soviet army, it might be nothing but a collection of lies that everybody told to everybody else, all in support of the biggest lie of all: that communism was working.

In his 1996 interview, former Director of the NRO Martin Faga made another novel argument that the intelligence community now takes for granted: satellites could not do everything. What the military needed was drones that could go into areas where the United States did not necessarily have total control of the skies and then operate there for a long period of time. By the early 2000s, the era of “persistent surveillance” by unmanned drones like the Air Force’s Predator became common. But whereas satellites cannot hover over a single location for a long time, aircraft cannot go over what the intelligence community still refers to as “denied territory.” To do the intelligence mission, you need both, and even then there will be many unknowns.

The bus-sized HEXAGON reconnaissance satellite was the largest intelligence satellite launched by the United States in the 1970s. Throughout that decade and into the early 1980s, the National Reconnaissance Office evaluated launching, and recovering HEXAGONs using the space shuttle. (credit: NRO)

The big bird’s swan song

Although it has been gone for decades, the HEXAGON’s capabilities have never been replicated. The former NRO official believes that its retirement was as much budgetary as it was technological. “When the stories are written I believe they will show that there wasn’t enough budget or will to fund anything other than the electro-optical system as they wanted it to evolve. Any film system would be competition for the funds and a constant reminder of the electro-optical system’s limitations,” he explained. “It was time to force the space reconnaissance community to the next plateau whether they wanted to or not. The electro-optical system did offer the near real-time capability that the film systems lacked, but at the expense of a lot of collection area.”

The former official noted that in some ways this was ironic: in the 1950s and ’60s, when the CIA and Air Force were jockeying for their respective positions on the first imaging satellites, the Air Force pushed an almost real-time imaging system named SAMOS to help with the Air Force’s attack warning and war fighting mission. In contrast, the CIA advocated a film return system named CORONA to do their National Intelligence Estimate assessment mission. “Then in the ’80s the roles reversed, with the CIA pushing the electro-optical near real-time system and the Air Force pushing the film-return systems. Organizational missions hadn’t changed, but the implementation approaches were reversed. Each was pushing the technology they had.”

Today, 25 years after that history was written, and over three decades since the last successful HEXAGON mission, HEXAGON’s unique collection capability has almost surely still not been matched

In 1992, the NRO produced a secret history of the HEXAGON program that was finally declassified in 2011. It summarized the program’s capabilities by stating: “During its 13-year life, HEXAGON provided a unique collection capability which may never again be achieved by US imagery satellites. Its ability to cover thousands of square nautical miles with contiguous, cloud-free, high-resolution imagery in a single operation provided US intelligence users and mapping, charting, and geodesy (MC&G) organizations with vast amounts of nearly simultaneous contiguous coverage. Order-of-battle information across entire Soviet military districts could be achieved in a short timeframe. Sino-Soviet military tactics could be studied and determined by analyzing imagery of Warsaw Pact, Soviet, and Chinese large-scale exercises. HEXAGON provided the best MC&G support ever furnished to the user community.”

Today, 25 years after that history was written, and over three decades since the last successful HEXAGON mission, HEXAGON’s unique collection capability has almost surely still not been matched. Even today, the story of the HEXAGON—and the space shuttle—is still being written, but one thing is increasingly clear: HEXAGON was unique, incredibly powerful, and missed by the intelligence community and the military.


These three links provide more history to the Hexagon story.

Part 1:

Part 2:

Part 3



Designers and Draftsmen in the 1970's

This is a 1970’s photo of some of the Danbury PE design and drafting personnel who created design drawings for the Hexagon program. CAD did not exist in those days and everyone used large drafting boards. The drawings were created in pencil on transparent drafting paper and if a change had to be made, erased and corrected or totally re-drawn. When finished, the tracing paper was put through a large machine containing ammonia that then developed the final blueprint on a special paper. The ammonia smell, especially for large drawings, lasted for weeks.

As best I remember the names from left to right were: Al Gasper, Tom O’Mara (manager), Ray Schneider, unknown, unknown, Don Boughton (bearded and later Mayor of Danbury), unknown, Tom Hayden, Jack Gearin, Juneau Voller, Gil Gonzalez and Ralph Forsberg. Unfortunately I do not have a photo of many other designers who worked on the program. They all were skilled and made many significant contributions.  

I ask that any Perkin-Elmer former employee or still currently working please send me any photos, memories or stories about your days at PE. My email is

Design &    Drafting, JPEG .jpg

Two Hexagonians

Is this going too far? Here are Jon Aspinwall and Phil Pressel showing their shirts and license plates as alumni and in honor of the job of a lifetime. 



Phil Pressel .jpg


 Clever Jon Aspinwall put the photos together with Hexagon hats. 

Clever Jon Aspinwall put the photos together with Hexagon hats.