Russians actually built a "Doomsday Machine."

[weemadando]

We cannot allow a mineshaft gap.

The Soviets Built a Doomsday Machine. It’s Still Working.

* By Nicholas Thompson Email Author
* September 22, 2009 |
* 4:20 pm |
* Categories: Russia, Weapons and Ammo
*

In the 1980s, the Soviets built a Doomsday Machine — and it’s still active. I learned about it while researching my new book, The Hawk and the Dove: Paul Nitze, George Kennan, and the History of the Cold War. It’s also the subject of a story in the current issue of WIRED.

The system would have allowed the USSR to respond in semi-automatic fashion to an American nuclear strike — even one that had knocked out the Kremlin and the Defense Ministries. And one obvious question that people ask when they learn about this system is “Why didn’t the Soviets tell us?” As we all know from Dr. Strangelove, the point of a Doomsday Machine is to convince the other side in a conflict that attack is futile.

The answer to that question is the most interesting thing I learned in my research. Yes, the Soviets were extremely secretive; and, yes, they were worried that, if they told us, we could disable it. But the more interesting reason is that they also built the system to deter themselves. Here’s the relevant passage from the story:

By guaranteeing that Moscow could hit back, Perimeter was actually designed to keep an overeager Soviet military or civilian leader from launching prematurely during a crisis. The point, says [former Soviet space official Alexander] Zhelenyakov says, was “to cool down all these hotheads and extremists. No matter what was going to happen, there still would be revenge. Those who attack us will be punished.”

Perimeter also bought the Soviets time. After the US installed deadly accurate Pershing II missiles on German bases in December 1983, Kremlin military planners assumed they would have only 10 to 15 minutes from the moment radar picked up an attack until impact. Given the paranoia of the era, it is not unimaginable that a malfunctioning radar, a flock of geese that looked like an incoming warhead, or a misinterpreted American war exercise could have triggered a catastrophe. Indeed, all these events actually happened at some point. If they had happened at the same time, armageddon might have ensued.

Perimeter solved that problem. If Soviet radar picked up an ominous but ambiguous signal, the leaders could turn on Perimeter and wait. If it turned out to be geese, they could relax and Perimeter would stand down. Confirming actual detonations on Soviet soil, after all, is far easier than confirming distant launches. “That is why we have the system,” says Valery Yarynich, one of the system’s designers. “To avoid a tragic mistake. “

[Samuel]

How did they expect it to work? I was under the impression this was impossible, and given it was small enough to be a secret...

[The Duchess of Zeon]

Soviet leadership enters launch codes and activates the distributed system. System holds launch codes until it detects nuclear detonations within Soviet territory, and then automatically releases the launch codes from presumably massively reduntant sites in a massively redundant communications infrastructure. It's actually extremely simple.

[Vympel]

Yeah, I knew about Perimeter - but it's hardly a 'Doomsday Machine' in the Dr. Strangelove sense. The one in Dr. Strangelove was designed to destroy the world, Perimeter is more of a 'dead hand' (if that's the right expression) to ensure Soviet retaliation even in the event of a decapitating strike.

Nowadays it ensures Russian retaliation.

[Count Chocula]

What Duchess said. It's a high-tech deadman's switch applied to a nuclear arsenal, and an approach that the US, with our positive control protocols over nuclear initiations, would never take. That said, I'd be surprised if we did not know about it. And use that knowledge as a spur to THAAD. The Nike/Zeus, Safeguard, Sentinel, and SDI programs weren't developed in an intelligence vacuum.

[Sarevok]

So its more like a soviet version of Skynet instead of a Strangelove esque Doomsday device ?

[weemadando]

I do like the fact that it was also designed as a way of preventing leaders from lauching pre-emptively, by assuring them that retaliation would occur if the enemy escalated first.

[The Duchess of Zeon]

So its more like a soviet version of Skynet instead of a Strangelove esque Doomsday device ?

More or less, though the automation would be ridiculously simple and further from AI than a modern day laptop is.

[Lord Relvenous]

I do like the fact that it was also designed as a way of preventing leaders from lauching pre-emptively, by assuring them that retaliation would occur if the enemy escalated first.

I liked this myself. Good way to remove the "clock is ticking" stress of these situations.

[Lord of the Abyss]

Yeah, I knew about Perimeter - but it's hardly a 'Doomsday Machine' in the Dr. Strangelove sense. The one in Dr. Strangelove was designed to destroy the world, Perimeter is more of a 'dead hand' (if that's the right expression) to ensure Soviet retaliation even in the event of a decapitating strike.

"The Dead Hand" is the name I heard used for the system back when news of it first came out, shortly after the USSR fell. This is the first I've heard the name "Perimeter" given to it, but I haven't been paying close attention to news about Russia in recent years.

[Patrick Degan]

Such a mechanism is actually discussed in both the book and movie Fail-Safe: in the movie, Prof. Groteschele (Walter Matthau) speaks of a "doomsday system" involving missiles that would go into action weeks after an attack which the Soviets "must" know about and would "destroy an enemy even after that enemy has destroyed us". In the original novel, Groteschele's description is more detailed in speaking of "doomsday tapes" in the command-and-control computers. Now, if Eugene Burdick and Harvey Wheeler were merely plucking that out of thin air, they were making some pretty good guesses and of course never knew how close to reality they were when they wrote that scene in their novel.

[K. A. Pital]

A deadhand switch is a "doomsday machine"? Well, for some... Incidentally, Perimeter was much discussed a few years ago in the Russia media, after which it fell into obscurity again.

[Simon_Jester]

A deadhand switch is a "doomsday machine"? Well, for some... Incidentally, Perimeter was much discussed a few years ago in the Russia media, after which it fell into obscurity again.

It is in fact a machine that can potentially bring about doomsday even when no one specifically tells it to; as I understand it, the system controlled a lot of megatons. So the name "doomsday machine" is reasonably accurate, I think.

The chief difference between this system and the one portrayed in Dr. Strangelove is that the Perimeter system can be turned off. Which is significant, but only insofar as it is further proof that the Soviet Union was not controlled by dark-comedically incompetent morons.

[K. A. Pital]

No, the dead-hand switch cannot simply automatically bring doomsday. And if it detects mass nuclear detonations plus gets the command codes because human-manned RVSN posts by that time would determine they are under attack, well, presumably the doomsday has already occured, has it not? Acting with minimal human input is not the same as acting totally without such input. The command missile can initiate nuclear war without human input, but to launch it, Perimeter has to receive so many multiple confirmations, both from automatic systems and human-manned systems, that it hardly can err at all.

[The Duchess of Zeon]

Honestly it's a brilliant system, because it does remove the urgency; it essentially means that a decapitation attack cannot occur. At the slightest warning, Perimeter will be given the necessary codes to activate, and then the leadership does not have to worry about the result if they refuse to attack immediately, and instead try to de-escalation the situation knowing that if they abruptly vanish in a ball of nuclear fire, perimeter will activate and the nuclear retaliation will begin.

[Bluewolf]

What would it actually take for the US to hypothetically destroy or retard this machine? It said in the article that the Russians were afraid of the machine being stopped somehow by the United States.

[Ryan Thunder]

Missile interceptor systems in Europe and the Far East to intercept the missiles in the launch stage, plus BMD that can handle whatever's left over North America?

I think, anyway.

[Narkis]

But that wouldn't really disable the machine itself. It would neutralize the Soviet missiles, yes, but Perimeter would still do its job dutifully.

[Sea Skimmer]

What would it actually take for the US to hypothetically destroy or retard this machine? It said in the article that the Russians were afraid of the machine being stopped somehow by the United States.

As described the system relies on numerous dispersed computer systems, linked by land lines and multiple kinds of radio systems. The only way to stop it would be to know the location of every processing node (some of which may be mobile on trains, aircraft ect..) and destroy them. Some of the nodes are also within the deepest bunkers the world has ever known and which can withstand a direct hit from multiple nuclear weapons before they die. So any attack large enough to wipe out Perimeter would be so big by default that the Russians would simply fire every nuke they have already.

The only way that wouldn’t happen is if somehow the US achieved total strategic surprise, as well as having perfect intelligence, and had a B-2 bomber already circling not only every Perimeter center but also every normal command and control center (it accomplishes nothing to kill only one or the other) and had them all simultaneously drop nuclear weapons on a time on target attack that totally wipes out everything.

Not a very likely thing to occur, particularly as the Russians would certainly pay close attention to any B-2 bomber movements and do posses certain SAM and radar systems which could engage that aircraft. This would also be assuming the US had produced the planned 100+ run of B-2s not a mere 21 of the things as the Russians easily have that many command and control bunkers. Plus even if all normal command and control bunkers die, the Russians still have bunkers in the silo farms, and SSBN commands and mobile missile systems all of which may gain a default permission to fire by the destruction of Perimeter itself. So a strategy like this could probably never work. Certainly it’d be far too risky to be attempted, even if a nuclear war somehow could not be avoided.

[Simon_Jester]

No, the dead-hand switch cannot simply automatically bring doomsday. And if it detects mass nuclear detonations plus gets the command codes because human-manned RVSN posts by that time would determine they are under attack, well, presumably the doomsday has already occured, has it not? Acting with minimal human input is not the same as acting totally without such input. The command missile can initiate nuclear war without human input, but to launch it, Perimeter has to receive so many multiple confirmations, both from automatic systems and human-manned systems, that it hardly can err at all.

Sorry. I misunderstood. I retract my previous statement.

What would it actually take for the US to hypothetically destroy or retard this machine? It said in the article that the Russians were afraid of the machine being stopped somehow by the United States.

Tongue in cheek: the most obvious way to neutralize this machine is to not fire nuclear missiles at the Soviet Union. Which, I suppose, was the point.

[Bilbo]

What would it actually take for the US to hypothetically destroy or retard this machine? It said in the article that the Russians were afraid of the machine being stopped somehow by the United States.

As described the system relies on numerous dispersed computer systems, linked by land lines and multiple kinds of radio systems. The only way to stop it would be to know the location of every processing node (some of which may be mobile on trains, aircraft ect..) and destroy them. Some of the nodes are also within the deepest bunkers the world has ever known and which can withstand a direct hit from multiple nuclear weapons before they die. So any attack large enough to wipe out Perimeter would be so big by default that the Russians would simply fire every nuke they have already.

The only way that wouldn’t happen is if somehow the US achieved total strategic surprise, as well as having perfect intelligence, and had a B-2 bomber already circling not only every Perimeter center but also every normal command and control center (it accomplishes nothing to kill only one or the other) and had them all simultaneously drop nuclear weapons on a time on target attack that totally wipes out everything.

Not a very likely thing to occur, particularly as the Russians would certainly pay close attention to any B-2 bomber movements and do posses certain SAM and radar systems which could engage that aircraft. This would also be assuming the US had produced the planned 100+ run of B-2s not a mere 21 of the things as the Russians easily have that many command and control bunkers. Plus even if all normal command and control bunkers die, the Russians still have bunkers in the silo farms, and SSBN commands and mobile missile systems all of which may gain a default permission to fire by the destruction of Perimeter itself. So a strategy like this could probably never work. Certainly it’d be far too risky to be attempted, even if a nuclear war somehow could not be avoided.

The system did have its holes. First a complete surprise attack with little warning could take out the command people before the system can be turned on. This is not very likely since having a system like this in place means its going to be activated any time there is heightened tension and what are the odds that a strike is launched just out of the clear blue.

Another hole is that the main article I read mentioned that the first missiles fired were sort of command missiles that radioed down instructions to other missiles. Being the first launched I would assume (though I could be completely wrong) that they would be the first targeted by any ABM systems in place. If those command missiles are destroyed early enough then what effect would that have on the Perimeter system?

Overall though I like the system. Its not so much the petulent revenge system as it is a stay calm even if they hit first we will still hit back calming feeling. Removes the whole "we have to fire now before they kill our ability to fire" fear.

[Sea Skimmer]

The system did have its holes. First a complete surprise attack with little warning could take out the command people before the system can be turned on.

The whole point of the system is that it provides redundant command and control links to ensure that a very large number of people could order a nuclear strike. Its Perimeters job to ensure, that the senior surviving general or political official is the one who gets to make that call. You have to destroy the entire system to prevent this, hardly a hole as such a massive attack could not achieve complete surprise in the first place.

Another hole is that the main article I read mentioned that the first missiles fired were sort of command missiles that radioed down instructions to other missiles. Being the first launched I would assume (though I could be completely wrong) that they would be the first targeted by any ABM systems in place. If those command missiles are destroyed early enough then what effect would that have on the Perimeter system?

The US had radio command broadcasting missiles like that too, converted Minutemen. They are only one of several systems which can authorize individual silo farms to fire. ABM is totally irrelevant unless the US had an airborne laser plane circling each launch silo. The need only rise up a few hundred miles in ordered to get radio LOS to the entire country, not fly a 6,000 mile long ballistic arc to the US.