How Hot is Ukraine Gonna Get?

[danfrodo]

25 minutes ago, Elmar Bijlsma said:

Heya Charles! Long time no see. How is the brain in a jar doing?

I particularly enjoy the Kherson part as it was so obviously a trap and the Russians fell for it anyway. Only to find it was both trap AND decoy. The whole thing is just a layer cake of cunning plans.

It's an interesting kind of trap.  Because it requires the opponent to be the stupidest idiot on the planet.  Requirement was not just met but idiot reinforced the stupid by shoving more of his scare resources, and some of his best, into the trap.  I am still dumbfounded by this.

[Beleg85]

Dmitri guy publishes talk by Ukrainian defence analyst, worth to read:

https://wartranslated.com/update-from-ukrainian-military-expert-oleg-zhdanov-12th-september/

 

It seems that (as it could be expected) news about Russians surrendering in Kherson were slightly "premature" or very local. I would advice caution as to assessing progress in Kherson as well, including map by Wolski. Too much noise to draw any conclusions now.

[Haiduk]

19 minutes ago, Beleg85 said:

Dmitri guy publishes talk by Ukrainian defence analyst, worth to read:

Zhdanov is the same sort "experts" like Arestovich, but if auditory of Arestovich are mostly people, who nothing understand in warfare, that Zhdanov is for more competent audience. Anyway both are supported by President's Office. Both have to radiate optimism to society. Zhdanov is retired colonel (retired in 2007), but his military analytic attempts and prognosіs during ATO were complete bull s...t. If you need real good situation assesment - Mashovets is the best since 2014.

Edited September 12, 2022 by Haiduk

[Grigb]

Wagnerite told Christo Grozed that there will be new stage in October.

 

Could it be RU is planning to make offensive in October in Zaporozhie direction? Let's say they decided to minimize scope of war to just "liberate of DPR". They actually said it.

Quote

The Ministry of Defense of the Russian Federation stated that in order to achieve the goals of its liberation of Donbass, it was decided to regroup Russian troops in the areas of Balakleya and Izyum to increase efforts in the Donetsk direction.

The only question is what is going to happen first? New RU offensive or RU Kherson collapse. 

[sburke]

1 minute ago, Grigb said:

Wagnerite told Christo Grozed that there will be new stage in October.

 

Could it be RU is planning to make offensive in October in Zaporozhie direction? Let's say they decided to minimize scope of war to just "liberate of DPR". They actually said it.

The only question is what is going to happen first? New RU offensive or RU Kherson collapse. 

I'm not holding my breath on another Russian offensive.  They have far too much damage control to deal with.  That being said I think it would be welcome to see them do so.  It'll be easier than digging the rats from their holes.

[danfrodo]

He says it's all because of corruption and incompetence in the army ("the mercenaries were scammed of their pay"). But is again very optimistic, says each government corporation and oligarch are told to set up their own PMC for "new stage in October".  

From Grigb post above.  Heck, I want in on this also!  I'll form a PMC and it'll have thousands soldiers in it, just look at all the signed contracts I have!  So as soon as you send 3 months pay we'll be headed for the front.  Here's my swiss banking info for the deposit.  talk about opening the corruption floodgates

[A Canadian Cat]

20 hours ago, Battlefront.com said:

I haven't a clue how to stop an AI drone swarm

Something like this would be where I'd start:

https://thenextweb.com/news/researchers-tricked-ai-ignoring-stop-signs-using-cheap-projector

The issue with AI is that as amazing as it is, AI does not understand anything, not a thing. It's all about pattern recognition and algorithms to react to said patterns. Learn to disrupt or confuse that and the AI no longer does what the author / owner wants.

Edited September 13, 2022 by IanL
Typo

[Centurian52]

On 9/10/2022 at 9:04 PM, Calamine Waffles said:

How do you achieve said breakthroughs using purely infantry and artillery?

Through a willingness to tolerate higher casualties. It's not that it's impossible to fight a war without tanks. It's just really inadvisable. I'm with you that tanks are critical to modern warfare (and UGVs really do just sound like next-gen tanks to me), but when you are arguing fervently for or against something it can be easy to paint yourself into a corner by making absolute statements. The trouble with absolute statements is that your opponent only needs to find a single counter-example to defeat your argument. Breakthroughs can be achieved with infantry and artillery only, and no tanks. They are just harder, require a higher tolerance for casualties, and have a lower chance of success.

[poesel]

Sorry

[danfrodo]

4 minutes ago, IanL said:

Something like this would be where I'd start:

https://thenextweb.com/news/researchers-tricked-ai-ignoring-stop-signs-using-cheap-projector

The issue with AI is that as amazing as it is, AI does not understand anything, not a thing. It's all about pattern recognition and algorithms to react to siad patterns. Learn to disrupt or confuse that and the AI no longer does what the author/ owner wants.

you are wrong, wrong, wrong.  AI/machine learning is a magic rainbow pony fairy princess that can predict anything given any small amount of data, no matter how corrupt that data.  Once trained, it can then predict anything outside the small, bad data samples it learned from.  You need to get egykated.  

[Grigb]

RU Nats are working to resolve RU Counter-battery problem. Expect that type of drones to appear in several months,

Quote

One of the big problems we have is the lack of a suitable process for counter—battery warfare against guns located 20 km and further from the contact line. That is why, in particular, it is impossible to suppress everything that hits Donetsk. There are counter-battery warfare radars ("Zoos") with an effective radius of up to 30 km, but they, in turn, are vulnerable themselves, they are easily found and can be attacked with PGMs. There are also acoustic reconnaissance systems — a several pieces for the entire [Invasion] grouping. You can't detect them at the time of operation, but they have a minus — reduced accuracy; they give out a very approximate circle on the map from where the launch could have been carried out.

The smart solution to the problem could be a reconnaissance and strike complex assembled literally on cheap, consisting of an acoustic complex and a relatively long-range drone (not a Mavic, but also not an Orlan, which is a fairly expensive and rare bird, but a relatively simple aircraft or VTOL on batteries, such as geoscan drones),

which should promptly launch into this circle with a flight task [of flying by] "the map" and have completely disabled signal reception, so as not to be affected by enemy electronic warfare. It flies, shoots and transmits the picture, after that the exact coordinates of the target are determined and given to the gunners.

The question is how to assemble a kind of mini-working group to develop and test such a solution. Now [we have] specialists in acoustic reconnaissance separately, then, of course, drone pilots, well, there are gunners too. There is no assembly point. It would be good if someone make an agile team that would tie it all into one [project]

 

[sburke]

13 minutes ago, IanL said:

The issue with Sburke is that as amazing as he is, Sburke does not understand anything, not a thing. It's all about pattern recognition and algorithms to react to said patterns. Learn to disrupt or confuse that and Sburke no longer does what the owner wants.

My wife edited this

[riptides]

2 minutes ago, danfrodo said:

you are wrong, wrong, wrong.  AI/machine learning is a magic rainbow pony fairy princess that can predict anything given any small amount of data, no matter how corrupt that data.  Once trained, it can then predict anything outside the small, bad data samples it learned from.  You need to get egykated.  

You are a few generations behind....

My algorithms figure out what they are looking at (data wise) when they load it and act accordingly......

Common recognition - a basic pillar of AI.

Just waiting on the load....waiting...waiting...

 

/S

[Grigb]

RU Belingcat partners CIT regarding RU army

Quote

Assessing the actions of the actually defeated 11th army corps and the huge amount of faulty equipment lost in the Izyum by the 1st tank army (including a large number of tanks), we decided to reduce our assessment of the effectiveness of Russian troops to "partually capable to defend" (on a scale from "capable to attack" to "unable to defend"). Previously, we were not sure that the problems we knew were so systemic, but now it turned out that the 11th Army Corps was completely incapable of defending the front line. It also became known that the organizational and staff structure of the 1st tank army was completely disrupted. This problem arose due to a large number of losses and prolonged participation in hostilities. The loss of two significant formations is a serious loss for the Russian army.

Now we doubt the success of any Russian offensive operations, but we admit that they still can be carried out.

 

Regarding the captured colonel @akd (Captured Lt. Col. Artem Khelemendik)

Quote

During the counter-offensive, Ukrainian forces captured a lieutenant colonel from the 18th motorized rifle division with the position of "head of the control/command post" of the military unit 90151. At the same time, he is the Deputy Chief of Staff. The head of the division's command post can serve as a valuable source of information about his unit for the Ukrainian military. And the capture of the control/command post indicates that this division has ceased to exist as an independent unit.

 

[Battlefront.com]

16 minutes ago, IanL said:

Something like this would be where I'd start:

https://thenextweb.com/news/researchers-tricked-ai-ignoring-stop-signs-using-cheap-projector

The issue with AI is that as amazing as it is, AI does not understand anything, not a thing. It's all about pattern recognition and algorithms to react to siad patterns. Learn to disrupt or confuse that and the AI no longer does what the author/ owner wants.

Lots and lots of limitations on AI, that's for sure.  The more complex a task you ask it to do, the more difficult it is to make AI capable of handling it.  I make wargames, so I know this to be true

AI works best when it isn't working alone.  In the case of drones, it has GPS to aide it's travels.  It can also have an internal guidance that uses flight data to know where it is and where it is going.  Not as precise as GPS. but it can work as a substitute for GPS failures.  All the AI needs to do is understand how to stay on a designated flight path and when to dive to the ground.  No pattern recognition, no optical systems, no remote signals.  It just goes where it is told and BOOM.

Fancier stuff like flying drones in complex environments does, of course, present the possibility of dazzling/confusing the input sensors.  Good AI and redundant sensors can make that more difficult, but it isn't as foolproof as plotting a clear path and letting navigation get it to target.

The defender has to take into take into account all possibilities.  That's expensive and perhaps even impractical.

Steve

[pavel.k]

There won't be a civil war in Russia. Instead there will be new era: Oligarchs Wars 2022 - 2036.

 

[Grigb]

39 minutes ago, Grigb said:

Wagnerite told Christo Grozed that there will be new stage in October.

 

Could it be RU is planning to make offensive in October in Zaporozhie direction? Let's say they decided to minimize scope of war to just "liberate of DPR". They actually said it.

The only question is what is going to happen first? New RU offensive or RU Kherson collapse. 

I forgot to mention - there are rumors that the status of war (currently it is SMO) is going to be changed. But there is no any details. So, it is possible that nothing will really change except some titles.

[UPDATE] just read one RU Nat post - Kremlin might want to change status of war to formally allow use of conscripts.  Good luck with that.

 

Edited September 12, 2022 by Grigb

[Battlefront.com]

1 minute ago, pavel.k said:

There won't be a civil war in Russia. Instead there will be new era: Oligarchs Wars 2022 - 2036.

 

I see this and I wonder if ANYBODY is advising Putin.  I mean, Holy crap this is really stupid.  The last thing Putin should want is a bunch of oligarchs, who are as a tribe unhappy with this war, not only raising and commanding battalions, but to also force them into funding them!  Wow.  This is a new level of stupidity.

Steve

[Maciej Zwolinski]

9 minutes ago, pavel.k said:

There won't be a civil war in Russia. Instead there will be new era: Oligarchs Wars 2022 - 2036.

I am waiting for some guy from Königsberg or Karelia to establish himself in Tuva or thereabouts and redo the adventures of v.Unberg-Sternberg as his reincarnation (the good Baron was a Buddist and strong believer in metaphysical stuff). My life would be complete.

[Centurian52]

On 9/10/2022 at 9:59 PM, Vanir Ausf B said:

I can't escape the feeling that some of the "lessons" we are seeing are artifacts of this war's peculiarities and may not be transferable to other conflicts, even near-peer. Is it that mass doesn't work or is it that you can't mass without denuding vast swaths of frontage, as the Russians are discovering to their chagrin.

Yeah, I suspect there is a significant amount of variability in warfare that is distinct from changes over time. Just as, in WW2, the fighting in North Africa was very different from fighting on the Eastern Front, and both were different from the fighting in the Pacific, I suspect that if another high intensity war between two different peer or near peer armies broke out in another part of the world it would look very different from this one in a lot of ways. It's not that this war has its peculiarities, so much as every war has peculiarities. Differences in objectives, scale, level of commitment, doctrine, force structure, and terrain may create a massive amount of variability even in wars fought in the same time period. Time period/technology obviously does make a big difference. If you reran WW2 with modern technology, but all other factors kept identical, it would still be a very different war. But I think it is far too simplistic to think of time period/technology as being the only thing that makes wars different.

For an obvious example, there is probably a comparable amount of difference between a modern land war in eastern Europe and a modern air/naval/amphibious war in the Pacific as there is between a modern land war in eastern Europe and a 1940s land war in eastern Europe, or between a modern air/naval/amphibious war in the Pacific and a 1940s air/naval/amphibious war in the Pacific.

So I don't think we should be talking about how modern war is different from war of decades past, as if modern war and war of decades past are homogeneous things, but about how specific types of war are different from their older counterparts. How is modern European ground war different from European ground war of decades past. How is modern counter-insurgency in desert/jungle/etc... different from counter-insurgency in desert/jungle/etc.. in decades past. How is modern air/naval/amphibious war around scattered island chains different from air/naval/amphibious war around scattered island chains of decades past. How is modern peer vs peer desert combat different from peer vs peer desert combat from decades past.

Edited September 12, 2022 by Centurian52

[Beleg85]

9 minutes ago, Battlefront.com said:

I see this and I wonder if ANYBODY is advising Putin.  I mean, Holy crap this is really stupid.  The last thing Putin should want is a bunch of oligarchs, who are as a tribe unhappy with this war, not only raising and commanding battalions, but to also force them into funding them!  Wow.  This is a new level of stupidity.

Steve

It probably depends who will lead them, I expect something more like common fund they pay money into that is then used by official state apparatus to attract mercenaries. Otherwise it would really be uterly stupid.

[pavel.k]

23 minutes ago, Battlefront.com said:

I see this and I wonder if ANYBODY is advising Putin.  I mean, Holy crap this is really stupid.  The last thing Putin should want is a bunch of oligarchs, who are as a tribe unhappy with this war, not only raising and commanding battalions, but to also force them into funding them!  Wow.  This is a new level of stupidity.

Steve

Peter the Great?

[Beleg85]

For people who like recon, it looks like proper first entry of Ukrainians into Kupyansk. Are near the bridge:

 

[Sojourner]

1 hour ago, Grigb said:

Could it be RU is planning to make offensive in October in Zaporozhie direction? Let's say they decided to minimize scope of war to just "liberate of DPR". They actually said it.

Or, could it be PMC is planning to make offensive on 25 October in Petrograd Saint Petersburg direction?

[akd]

27 minutes ago, Beleg85 said:

It probably depends who will lead them, I expect something more like common fund they pay money into that is then used by official state apparatus to attract mercenaries. Otherwise it would really be uterly stupid.

They already have access to centralized piles of cash to pay mercenaries.  They want the Oligarchs to leverage their own loyalty networks for “volunteers.”

Edited September 12, 2022 by akd