84 comments

[ 2.5 ms ] story [ 172 ms ] thread
Call it a hunch, but if the flares spew dirty and irregular, maybe there's an awareness that this sloppy spread of IR readings has benefits.
Would not surprise me. Soviet era designs and engineering seems to always having been more pragmatic, out of basic necessity.
The other way to read the article is to blame Soviet quality control. If we assume the variation in burn is random, it may mean overall better protection of your airborne resources, but you could not assess the protection of any individual aircraft. That sounds Soviet to me.
Like taking a vaccine. There will be a few who have a negative reaction but the positives greatly outweigh the negatives which makes it worth it for the greater good.
Usually you would want the flare to mimic the spectrum of your own engines. I don't know if by irregular they meant in spectrum or illumination level. There is not a lot of short term variation in a jet exhaust due to thermal mass and the need for thrust to keep maneuvering, but the aspect changes. So sensors are not assuming a straight tail chase.

I consider it much more likely that the Russians at some point obtained some AIM-9X seekers and tweaked their flares and/or engines to make it less effective.

The is also the concept of the 'war load' or 'war paint'. In normal operation planes will have their radar cross section increased and corner reflectors added, engine burn is tweaked, etc, anything that could be leaked or sensed remotely (when flying in international airspace, or from space, or covertly through radar backscatter). When you think things will get serious you pull out those chunks of superfluous steel, tweak the engine exhaust nozzle, load different firmware onto the missiles and the EW system. If actual combat occurs there may be different war-only RADAR wave forms.

Probably none of those are available to Syria (and the US has been building RADAR signatures of them for many months), except maybe an upgraded flare dispenser.

Interesting. Sloppy is random is simple is robust.

This reminds me of a story where the USA got its hands on a new MiG and when they opened it up found that it used some kind of antiquated technology. Some time later the USA found out that this was intentional because the older tech could withstand nuclear radiation whereas the new tech would instantly fail.

Was the Radar on the Mig-25 that a defector landed in Japan, they used Valves when all the US radars had moved almost entirely to transistors.

Fun history, the Russians demanded it back off the US/Japanese so the Japanese government let the US take it apart, photo and test the parts (as quickly as possible) then shipped it back in crates and billed the Russians for the shipping. Which iirc they never paid.

Thank you! I searched for more background and found another guess why the Soviets used valves instead of transistors:

"Unable to afford the thermal-resistance provided by a titanium frame, the MiG-25’s designers used vacuum tubes over transistor-based modern electronics — which made the planes resistant to high-temperatures and cheaper to maintain."[0]

[0] http://warisboring.com/one-soviet-defector-shoved-japan-into...

I think that detail about the billing USSR for the return shipping of the plane is propaganda meant to humanize the US/Japanese side and dehumanize the USSR side by showing that we have a sense of humor and they don't.

USSR could have replied by sending an invoice for the rental of the plane, reassembly of the components, and the intellectual property licensing for all of the technical data obtained. All of that would no doubt ought weigh the cost of crating and shipping the plane by orders of magnitude. Not that either side would ever have considered paying, of course.

In reality, some entirely secret concession could have been made that we might never find out about unless it was documented and is some day declassified.

Was the Radar on the Mig-25 that a defector landed in Japan

There's a similar story about the "sloppy" riveting work on the Mig-25. Turned out to have significant beneficial effect in the airflow.

The USSR was brilliant at doing more with less (advanced technology).

That was brought home to me in a documentary I watched about the Mig-29, the US/Nato avoided FOD (Foreign Object Damage) by scrupulously walking the runway/apron every day and cleaning up.

The USSR just put air intakes on the top for taxiing so crap didn't get sucked into the main inlets.

It was such a simple design and yet the advantages in a war situation operating from crappy debris strewn airfields where obvious, they had Mig-29's taxing over rough dirt berms and all sorts of crazy things (at least from a western viewpoint).

The AIM-9X uses an infrared imager which is why it is "all aspect" -- it is chasing a plane and not a hot spot per se. Dirty flares should not impact this too much in principle.

Given the age of the software in these missiles, the issue is likely relatively weak and brittle classifiers being used to separate noise from signal. It is not based on any kind of modern deep learning techniques.

Overfitting to the training data. Sounds vaguely familiar.
(comment deleted)
If you enjoy reading about electronic warfare, it is worth reading Steve Blank's "Secret History of Silicon Valley".

Here's the bit about moths [1]:

> John said, “Listen, you can hear the jammer.” The what? “The jammer,” he said, “Watch the moths.” It turns out the moths, through evolution, had developed their own electronic countermeasures to jam the bat radar. They had developed ultrasonic receivers and ultrasonic jammers and physical countermeasures. When they picked up the bat radar illuminating them by sensitive hairs on their antennas, they would send out their own little squirt of ultrasonics by rubbing their legs together, jam the bat radar, and then they would immediately take evasive action and dive to the left and right.

> Through Darwinian selection over millions of years, these moths had developed an entire electronic warfare, electronic countermeasures, electronic countercounter-measures suite, and here was a guy in 1973 in Thailand who was figuring this stuff out. To be honest, it was my first insight that there was really a bigger picture.

> So, John’s point was, “I keep trying to tell officers way above me that there’s probably a ton we could learn from watching these natural systems. What we’re doing in the air war over the North is just nothing more than something that’s been going on in nature for millions of years, but I can’t seem to get anybody’s attention.” (Thirty years later MIT would develop the Insect Lab and work on swarm behaviors for UAV’s and robotics.)

[1] https://steveblank.com/category/secret-history-of-silicon-va...

I've often wondered if a WW1 stringbag could fly right past modern air defenses without being detected or tracked.
Probably not, since modern air defenses are effective against small missiles and even mortar rounds. Here's a CIWS in Afghanistan shooting down incoming mortar rounds.[1] (Read the comments for more explanation from people who have been there.) The CIWS has radar which can track not just the target, but its own bullets. It makes quick adjustments until bullets and target are in the same place at the same time.

On one occasion, a US ally shot down a hobbyist-sized drone with a Patriot missile. Very expensive, but it hit the target.

[1] https://youtu.be/KsVUISS8oHs

An ababil UAV is larger than a hobbiest drone and the PAC5 apparently missed it and it was downed by a derby missile fired from a near by SPYDER battery.
Pac3 not 5 don't know why 5 got in there.
The comment sections (Alas1, this is Youtube after all), on the other hand, are depressing.
There was a team from a Large Defense Contractor at my local airfield recently with thermal-imaging equipment that seemed to be doing a good job of tracking light piston-engined aircraft, many of which are tube-and-fabric construction like WW1 machines.

They wouldn't say what they were testing but it seemed fairly obvious. The whole engine block glowed hot as did the uppersides exposed to the Sun, let alone the exhaust, at what I guessed to be several kilometres distance. Before we were shooed away :)

The engines also reflect radar quite nicely, even if the wood and fabric doesn't.
Not to mention the guy wires holding everything together.
WW1 was wood and fabric, not tube and fabric.
North Korea deploys canvas-on-wood Antonov An-2 biplanes as "low observable" stealth aircraft. Most of the radar energy goes right through it, there's very little metal to generate a return.

(Similarly, during the Battle of Britain the Hawker Hurricane was notable for the way German 20mm cannon shells passed through its doped cloth-on-tube airframe. No impact to trigger the fuzes.)

Another interesting property of the An-2 is the incredibly high cross-section of the aircraft means slow flight speeds and very high lift, the aircraft really doesn't have a stall speed due to spring-deployed slats that snap out as airspeed drops... but whatever you want to call the stall speed it's lower than 30 kts. So in a stiff headwind you can fly the aircraft onto the ground in full control at zero ground velocity, or actually while being pushed backwards by the wind.

Wouldn't be enough to deter a dedicated attacker at short range especially with a sufficiently high radar frequency (the engine and other metal parts become a bigger multiple of the wavelength), but would you be able to pick it up at 50-100 miles with longband search radar? especially at low altitude? Probably not. Especially random ex-Soviet crap like a lot of countries have.

Don't knock primitive tech, sometimes it works. The An-2 is archaic but it's a very competent bushplane/light cargo aircraft for squad-level paratroop or liason operations. And it's well-adapted for rustic operations like Russian/Canadian bush, or Korea.

https://en.wikipedia.org/wiki/Antonov_An-2

https://www.youtube.com/watch?v=A3tpV9vUTPo

http://www.smithsonianmag.com/smart-news/why-north-korea-usi...

I actually did a couple of skydive jumps out of old AN-2 planes. Interesting experience, but looking at them, warplane is not the word that comes to mind.
There were some incidents doing WWII where both the British and soviet ran successful raids using low and slow flying biplanes that was simply too low and slow for the motors guiding the anti aircraft guns of the day.

Doing the Taranto raid the low altitude of the Swordfish bombers caused the Italian navy to not just miss the planes but inflict collateral damage on themselves.

Today it's much harder to pull a similar stunt mostly due to night vision goggles/binoculars. But especially the US have a vulnerability as they have replaced proven effective systems based around cheap ammo and smart people, with systems based on dumb computers and expensive missiles.

The same thing, combined with daredevil pilots (or in this instance, navigator), enabled the Bismarck sinking.

Basically the Bismarck AA guns either had problem tracking them, or the ammo didn't do much.

And the torpedo that hit the Bismarck rudder was fired in part by the plane's navigator hanging over the side of the aircraft, and telling the pilot when to release to avoid the waves.

The Swordfish wasn't totally low-tech, they were equipped with radar which allowed them to find the Bismark at night and find their own carrier again afterwards.
(comment deleted)
I remember someone expressing surprise, years ago, that glass fiber gliders were detected by air traffic control radar. The most massive metal parts would be the undercarriage (single-wheel), but my guess is that the control cables, laid out in a cross at 90 degrees, would be the most reflective feature.
It doesn't take metal to reflect. Any change in dielectric constant causes a reflection, and that's amplified for any geometries that are multiples of 1/2 wavelength.
Indeed, but it is a matter of degree - glass fibre seems to be a common material for radomes. I made a cursory search for relevant figures, but didn't come up with specific figures.
Not just with airplanes! See the Millennium Challenge 2002 naval exercise:

https://en.wikipedia.org/wiki/Millennium_Challenge_2002

TL;DR: the Red team general, with Iran-equivalent forces at disposal, employed asymmetric tactics and WW2 equipment to sink an aircraft carrier, ten cruisers and five amphibious landing ships, estimated loss of 20 000 US military personell. Then they stopped the exercise because it was too embarrasing, refloated the Blue team ships and instructed the Red team to play by the instructions designed to make them lose.

I find it interesting the US jet was an FA-18, which entered service November 1983 (USN) is actually more than 30 years old as well.

https://en.m.wikipedia.org/wiki/McDonnell_Douglas_F/A-18_Hor...

This is equivalent to say a couple years old Ford F-150 is a truck from 1975. The branding is the same, the shape is similar, but it ain't the same plane.

Super Hornets are still being produced, upgraded and improved and will probably still be in the assembly line in 20 years. A 30 second Google didn't show the year of the F/A-18 in question, but I assume it is from the mid 2000 or something like it.

This was even more true for Soviet tech. The American B-52 are ancient, but the Soviet Bears were produced, upgraded, replaced and improved until the very end. Cost-plus programs like the F-35 with enormous upfront R&D are good for the manufacturer, suppliers and lobbied politicians, but iterative designs, like evolution, also just works.

The Tu-95 "Bear" is still is service:

https://en.wikipedia.org/wiki/Tupolev_Tu-95

Yes, but my point was more about the Soviet program of continuous improvement associated with it. According to your link, they stopped the assembly line at the collapse of the USSR. It doesn't prevent them from upgrading the existing airframes like for the B-52, but it isn't a true rolling replacements/improvements program anymore like the F/A-18 [almost] is.

Edit: Maybe the F-104 starfighter program would be another good example.

Worth noting that the Su-22 is actually just a variant of the Su-17 which first was produced in 1966.
Actually it was super hornet. It has so few common parts with the original, that many people consider it completely different airplane. The program costs we're also big enough to hint new design if I recall correctly.

The reason it is called "hornet" and why it shares so much external similarity is some kind of domestic propaganda reasons. To make people think that US Navy doesn't have modern equipment when they actually do.

The more commonly cited reason is that it was easier to get Congress to approve a budget for an improvement on an existing fighter than for a brand new fighter program.
And indeed there is much less risk in improving a good design over starting a new project from scratch. One wonders what we could have accomplished if we have invested the trillion dollars we are putting into the F-35 into say unmanned drone fighters, or batteries so that we dont have to defend the middle east.
Block III F/A-18E : FA-18 :: Core i7 : 386DX.
It's a 14-year old missile.[1] It's a variant of the Sidewinder missile from the 1950s. The Sidewinder has a good track record partly because it's so dumb. Classic versions are tail-chasers; the pilot must get on the opponent's tail before firing. Newer versions are supposed to be "all-aspect", but they still work better if they can see the hot end of a jet engine.[2]

[1] http://www.raytheon.com/capabilities/products/aim-9x/ [2] http://www.donhollway.com/foxtwo/

Obvious follow up question: soviet tech is easy to come by, at least the kind that's now used in dozens of countries. How is the missile not designed to be effective against the most likely target? Surely Raytheon must have a dozen versions of Russian and European flare dispensers and not just the US kind?
The answer: this is war propaganda. There are no real facts here, only half-truths and mis-direction.
What are you suggesting? That the Su-22 wasn't actually shot down? That the AIM-9X was actually successful, and the AIM-120 wasn't used?

You can't just say "blah blah propaganda, blah blah misdirection" every time you see a news article from a war zone.

Read what I said: this article may be full of misdirection and half truths.
No, you clearly did not say that the article may be full of misdirection. You very emphatically stated that the article certainly contained misdirection, but did not deign to tell us mere mortals what that misdirection might be:

"The answer: this is war propaganda. There are no real facts here, only half-truths and mis-direction." [Emphasis mine.]

Calling everything around you propaganda may seem like an easy way to look smarter than everyone else, but it is a painfully transparent move when you can't back it up with substance. It's even worse when you try to pretend you weren't doing it.

Okay, so the purpose of this article is what, exactly? To highlight the technological prowess of US forces when compared to their rivals?

And you think this doesn't involve any propaganda at all?

That's a fairly obvious dodge.

I would assume that the primary purpose of the article is to generate advertising and subscription revenue.

(comment deleted)
This is why Japan and S. Korea, as well as many EU countries are developing their own military weapons. They can't rely on US products being up to spec or even close to what is advertised. Instead the US resorts to strong-arming and bully tactics as they did with Saudi Arabia to sell their poorly developed shit.
Cancer on mobile. Kept sending me to the App Store to install various spammy apps.
Ah, but the modern AIM-9X is not so modern anymore. AI will be able to decrease false tracking.
still waiting for the answer, how was it able to do it?
The AIM-9X uses IR signature and it was trained to ignore US type flares which have a very consistent burn between different batches. Soviet era decoy flares are very irregular and for some reason the missile IR targetting prefers some of the Soviet flares to the Su-22 engine.
That's in the article, but it's pure speculation as applied to this particular event. It's possible the pilot screwed up or the missile just failed.
This is a perfect example where machine learning will overcome the issues of variability in burn times and intensity.

I think ML will have a great impact in warfare.

ML isn't going to help if all your training data comes from friendly flares.
You get your data when you fight an enemy airplane.
I think ML will have a great impact in warfare.

Here's a funny story, ML people like to talk about the ROC curve and the AUC as a measure of how well a machine learning experiment worked on the test data. Do you know where the term came from? WW2 radar engineers.

Also consider that bread-and-butter ML techniques like logistic regression and k-means clustering have been around for decades.

Out of curiosity, anyone know the overall expenses for an AIM-9X and AIM-120, vs. an SU-22?
Wikipedia lists: AIM-120 400 000$ and AIM-9X 600 000$.

Which is very surprising, typically short range missiles are cheaper. Apparently AIM-9X has gone through recent face-lift that still affects prices.

Su-22 is harder to find, but comparable aircraft SEPECAT jaguar goes for about 8 million.

Check wikipedia again

an AIM120C is 300-400k

an AIM120D is 1.7M (I think it was listed as 1.2M last I checked this page)

According to this http://www.f-16.net/f-16_armament_article1.html a sidewinder in 1999 had a unit cost of 55k-85k which would be 9L,M, or P most likely. 9X did enter service until 2003

These toys, much like the planes carrying them get progressively more expensive in newer version.

But the cost ratio stay much the same. You get 4-6 AIM-9 for the cost of one AIM 120

This unfortunately seems to again validate John Boyd's sentiment that a low tech, highly maneuverable, simple and reliable single purpose fighter with a simple gun is a better investment for actually winning a war than incredibly expensive high-tech planes tested by people with clip boards.

Edit: Between 1 F35 for $85 million or 5 F16 for $73 million total my money would most certainly be on the F16s.

It validates that the US military needs to test against other countries flares.
And then they make new flares without telling us.
Boyd hated the F-15. Funny how the F-15 is undefeated, and if you look at how it has fared in Israeli hands, it's one of the most exceptional fighters when it comes to actual combat...

And you seem to be implying that numerous SU-22s would have fared well against the Super Hornet? You could put 10 Fitters up against the Bug, and it'd still win before going winchester.

The super hornet typically only carries 6 air to air missiles. While SU-22 can carry the most modern short range aa missiles.

Hornet would not win with gun only against 4 missile armed fitters. If the missiles are modern. In air warfare numbers have always mattered and always will.

Iraqis thought the same thing, as did most of the Arab world when fighting the Israelis. While Stalin was partially right (quantity has a quality all its own), the most valuable factor in determining the outcome of air combat is pilot skill. And even if the Fitter pilots were as good as the Hornet pilot, they don't have as good ESSM/ECM systems, nor as good AWACs providing support.

The Bug doesn't even typically carry 6 AAMs. The one that did the shootdown probably only had 3-4 depending on its mission tasking. They have to worry about bring back weight when landing on their carrier. But for our hypothetical, the Bug can carry 12 missiles without a lot of trouble (1). It might not have much range, and the drag would be a bitch, but the Fitters would go down in flames...

1. http://runway.cz/img/zbr/aim120_2.jpg

Nonsense.

All A2A kills since 1999 have been with missiles. The F-16 have very preciously few A2A kills using guns.

Reminds me how su-24 completely shutdown a sophisticated Aegis defence system on US destroyer in a Black sea recently. The crew was so demoralized that they all quit upon return to the shore. http://m.huffpost.com/us/entry/11015434

I bet it was some old Soviet tech from the sixties.

This is fake news.
Are we certain that the first missile actually missed or could this just be America saying, "Oh, gee whiz, our missiles sure are terrible. No need for anyone to upgrade their countermeasures."
Obvious follow up question: soviet tech is easy to come by, at least the kind that's now used in dozens of countries. How is the missile not designed to be effective against the most likely target? Surely Raytheon must have a dozen versions of Russian and European flare dispensers and not just the US kind?
No mention, of course, that the American pilot was actually protecting the terrorists(even Michael Flynn talked about US financing the terrorists in Syria before he got sacked) the SU-22 was targeting but this is a different story here( here we are talking about geeky technical things like why the exceptional missile from an exceptional country was not so successful)
What acts of terror did those commit? Oh, none? So why are they terrorists and not freedom fighters?

How come "anti-establishment" doesn't mean anti-Russian/Syrian-dictatorship?

I just mention what your ex defence minister has told the public. Period. And there are no freedom fighters in Syria. Just missionaries on washington's payroll. http://www.informationclearinghouse.info/47311.htm
...in 2012-13 the Obama administration helped its Sunni allies Turkey, Saudi Arabia, and Qatar provide arms to Syrian and non-Syrian armed groups to force President Bashar al-Assad out of power. And in 2013 the administration began to provide arms to what the CIA judged to be “relatively moderate” anti-Assad groups—meaning they incorporated various degrees of Islamic extremism.
I'm not even on the same continent as the USA, and I don't reckon belonging to or owning any secretaries of defense.

That's certainly a nice collection of links, yet don't seem to answer my question at all.

Also, claiming there are not freedom fighters in Syria is denying the whole civil uprising in 2011, and all of the utmost brutality and horrors perpetrated by the Assad regime against those protesters.

I'm not advocating funding terrorist groups, nor do I think the CIA and the Pentagon did a great job. (The whole intel/defense community is in shambles. Too much internal infighting, too few people taking responsibility, too much politics, funding issues, too much corruption because too much high-tech gadgets that must be maintained, produced, supported. You know.) But your worldview is overly simplistic if you think the story is that "a US plane was protecting terrorists". Those fighting on the ground probably have a very flawed sense of religion and what's right, how the world should be, but they're are currently fighting a fucking dictator. And that's a good thing.

The situaion around Raqqah (where Tabqah is) is basically the Rojava trying to purge ISIS from Raqqah while keeping Assad at bay at Tabqah.

see https://en.wikipedia.org/wiki/Rojava for who these folks are.