kluge

In theory, this should be pretty easy as image classification is well known and widely studied.
Acquiring the data is a different story. Free data sources like FIRMS don't provide enough resolution to identify individual positions in any meaningful way. Commercial data sources have enough spatial resolution to identify vehicles, but generally don't have the temporal resolution to continually scan locations and doing so every hour is sure to rack up an eye popping bill. Government satellites definitely have the resolution, but that information is classified and its unclear if the US offers a direct data pipeline to Ukraine.

Not just good quality data. Good quality data that has been accurately labeled, and often in massive volumes and of high diversity.  Depending on your system, that can be very difficult to get - I do some optical systems where we're almost guaranteed to have low signal to noise (always working at the limit of our resolution) and are likely to run into things that are outside the training set, and we spend huge amounts of effort on both improving the repeatability of the acquisition (relatively easy), and automating the development of training data, which sometimes involves a lot of bootstrapping.  You think watching the scanners at TSA checkpoints is painful? Have I got a job for you...

You know it, babe 😙

 

I am confused.  Are we just talking about the Orlan?  Or the entire future of UAS?  I suspect the Orlan will be seen as a bi-plane by 1945 in a few years.

I don't think "knowing where the enemy is" is a problem if this war is a demonstration.  The ability to push small lethal drone swarms around on the battlefield only need to happen as fast or faster than your opponent can drive...or launch their own swarms.  Many ways to do this because each individual system is so light.  This means one can drop kill boxes at range and essentially deny an area.
As to sensors and processors.  Well we will have to see won't we.  Being able to fly through a bamboo forest autonomously looks pretty sophisticated to me and we are only getting started.  I honestly only see range and endurance as a local tactical limitation.  Operationally they can be packaged and projected at very long ranges by any number of systems.  A HIMARS cargo shell could carry a hundred of these small systems and deliver them 100kms away.  In place the little buggers could just sit on the ground until a target comes by and within range.  Trading off who pops up to take a peek or linked to other sensors that came with them (pigeon hearts anyone?).  5 fly up and overwhelm the target.  The other 95 stay in the grass.  I only need the little bugger to fly maybe a km.  Give them solar panels and they can recharge in place.
That is a major military problem. 

Just use dinosaurs, solves the problem. Orlan-10 and similar have loiter time in excess of 12 hours. Don’t be like Amazon’s drone program and insist on electric-only.
Yes, and no. Orlan-10 is pretty speed at 110mph. However, the current and next generation of these things aren’t maneuverable (especially the pusher prop ones), and have no capability to detect something targetting them.
Detecting, targetting and taking down these things without spending $1m is the real problem. You almost need something like an Orlan-10 itself that can loiter for a while and cover say a 40km square area, with some sort of micro missile optimized for cost and small targets, and a series of spectrum analyzers on the ground listening for control signals that are linked together.
Your air-to-air mini missile is not going to be expensive as far as real missiles go, but can you keep it under $50k? It needs to go in excess of 200mph (which has been done), and have 10km range, and have some sort of sensor head that can track the target. So yeah, that last one is the problem. Optical might not do it.
 
A book I recommend to everyone is “Sidewinder: Creative Missile Development” about how they developed Sidewinder on the cheapo.

A lot of incentive to solve the energy density problem of batteries:
https://www.technologyreview.com/2023/01/04/1066141/whats-next-for-batteries/?gad=1&gclid=CjwKCAjws9ipBhB1EiwAccEi1BVtyUUPriIA8d0KP2cg84dW41z0YN0FUllsukBpNjBKoZWhm2I20RoCHLoQAvD_BwE
This is much bigger than warfare and has eye watering levels of investment.  Time is distance.  So as loitering times lengthen, ranges are also going to stretch.  Time is options.  As loitering times increase so do the range of possible missions.  UGVs are waiting in the wings on this one too.
You best start believing in RMAs, Ms Turner…because you are in one.

Not to mention current loiter times. With half hour endurance a unit is looking at needing 48 small drones to keep a single unit in the air doing CAP 24/7. More if you count lost time for takeoff and recovery.

My guess it is the targeting.  Tracking, locking and hunting a small highly manoeuvrable object is a pretty high bar.  Basically trying to hit a magic bullet with a magic bullet.  Just because we have a similar system doesn’t mean we have the targeting capability to hit another drone.  And then one has to do it en masse. I have zero doubts that there are people working on this furiously.
This will set up an air power below 2000 feet arms race.  Counter counter UAS systems will then be developed.  I think a lot of people are starting to get that this is not War + Unmanned.  It is Unmanned Warfare.

I am highly skeptical about direct energy C-UAS, let alone C-RAM.  It may be technically feasible for static point defence systems one can plug into a power grid.  But for mobile defence I think the energy bill is just too high.  Further one would need thousands of these systems (which is what GD wants) to deal with the levels of systems we see in this war, let alone swarms likely in the next one.  These things will be 1) highly visible as that much energy being pumped into the environment is going to get detected…bit more juice than a pigeon heart, and 2) will be a prime target for all the systems that can see it.  We would likely lose them too fast to replace.
And then there is volume.  A direct energy weapon that can target and engage dozens or even hundreds of small tac UAS/loitering munitions at range would be a game changer for more than AD.  Such a weapon could target people and vehicles at ranges and rates that would change warfare forever if such a system existed.
No, this is typical western military industrial planning.  We have swarms of small UAS that costs a few hundred to a few thousand dollars each that we need to stop.  Of course we will develop and buy a laser- phased AD system that costs hundreds of millions and needs to be put on a big truck….makes perfect sense.
Or we take another UAS that costs a few hundred dollars and built it to kill other UAS.  

This is "Scorpion" - US-localized 120 mm Spanish "Alacran" SP-mortars. The system is mounted on Land Cruiser. According to Oryx two already were lost.
 

Hey look its Neuhof. 

Same pattern on both sides.  Try big, get hammered…go small.  Have western commentators go “silly [insert side], why are you going small?  Be more manoeuvry!”

Drones are the essential element of why the war works the way it does. We all know that. An effective technological solution that grounded one side drones would almost certainly result in a huge swing in the battlefield. The big question is does this solution exists, and can one side deploy it at scale faster than the other side can adapt. The other known tech shift that is just hanging out there waiting is much greater AI/autonomy/swarming. Could one side do that scale fast enough to matter.

Also, a lot of experts are yearning to see a large scale mechanised attack because that is what they think proper war looks like, and everything else must be an aberration or at best, coping with inadequate resources. I imagine a French or Prussian liaison officer looking with disdain at the Union attack at Cold Harbor in 1864 and shaking his head at the inability of the colonials to press home a bayonet charge or better coordinate attack columns etc.

From the Deep State Telegram Channel:
https://t.me/DeepStateEN/2646
🗡 The enemy lost at least 63 armored vehicles on the northern front of the Avdiivka direction
 
🛰 After analyzing the changes on satellite images, the loss of more than six dozen pieces of equipment was recorded. Almost all of the equipment was destroyed in the area of Krasnohorivka (A) and at the entrance to the settlement. As the classics would say - an illustration to the General Staff's reports.
 
🔥 These images show the entire "success" of the enemy's actions, which was mainly present in their information space. On the battlefield, despite their superiority on all counts, the katsaps lost the battle. Of course, it is difficult to estimate the enemy's losses in manpower. In this area alone, the number of dead ruzzians is measured in hundreds.
 
😑But there is also unpleasant news. This attack is yet another proof of the ineffectiveness of the armored fist tactic, when one side concentrates its equipment and tries to achieve success on the battlefield by using its superiority. Our destroyed columns near Mala Tokmachka, Verbove, Robotyne, Staromaiorske, Novodonetske and Rivne, as well as the enemy's destroyed columns near Vuhledar, Novoiehorivka, Makiivka, Andriivka, Marinka, and now Krasnohorivka (A) and Vodiane-Pervomaiske, are further proof of this thesis.
 
✈️ This is all a consequence of the growing effectiveness of aerial reconnaissance and artillery. The same cheap quadcopters equalize the capabilities of both sides. And so far, there is no technology that can provide an advantage. The goal of our military-industrial complex is to find the key to victory, otherwise history will simply repeat itself.
 
🖼 Full-resolution image.
 
📍 Location of the village - deepstatemap.live/en#13/48.2069/37.7256

Possibly most effective use of DPICM seen yet:
 

Let’s not make too long leaps here. (And I know you are just summarizing)  Russia may have reserves.  Or it may well have taken risks along line units as well.  Avdiivka was a tactical offensive, that failed - not D Day.  Russia is learning, there is no arguing that.  The question is: are they learning at a competitive rate compared to the UA?
UA takes ground continuously for months = “well ya but it isn’t fast enough”
RA does a post mortem twitch = “Holy Crap, it is the end of days!”

similar worry about this on the "War on the Rocks"
Some pointers from the episode:
Russian attack on Avdiivka is worrying. Russia has learned. The use of combined arms was on a completely different level than last year (artillery, air force, mechanized forces) and the scale was increased from companies to battalions. Even when it was a failure it was still a big improvement, as we have seen with Ukraine these types of operations are extremely hard to conduct even from competent and motivated forces. If this trend continues, there is cause for concern. Avdiivka shows that Russia feels it is in a strong position, at least proves that Russia is not on the ropes. Avdiivka risks becoming a new winter Bakhmut, i.e. a grinding, consuming battle. The culmination of Ukraine's offensive phase is around the corner. We will probably see the "last show" of this offensive within weeks. The likely goal is a breakthrough that would threaten Tokmak. The goals of the UKR offensives were not achieved. The political leadership announced the goals publicly at the beginning (minimum goal Tokmak, Bakhmut and Melitopol as maximalist obj.). Ukraine gets high marks for a good exchange ratio over the summer operation, especially taking into account the environment. Still, the substantial Russian losses seem to be sustainable for them. Now the eyes are on next year. Russia is going to have to conduct another major mobilization if it wants to "keep up the phase". Russian ammunition sources should not be underestimated. Iran, North Korea, etc. Russia has invested significantly in the war economy of these countries and its own. (note that this Ukrainian summer offensive was also carried out with E-Korean stocks) A significant challenge vs. Western production capacity. Next year, it will be challenging for the West to continue the same level of support as this summer, to enable Ukraine to keep the initiative. The political situation in the United States really becoming worrying. At the political level, complacency has spread in Europe and the United States that Russia is losing and cannot continue the war. The situation is the opposite, the West now has to make expensive decisions for a long war. Russia has resources and resilience as we have seen this year.

Another successful ATACMS strike, this time on a Luhansk airfield


Too bad we are seeing these strikes only now at the end of the Ukrainian offensive season. Probably these will be a big help for the awaited "last push" of the current southern campaign before culmination and mud season. 

Incidentally, the location on the left bank of the Dniper where this vehicle was blown up...
...is the exact same spot where Ukrainian forces were reportedly spotted recently.
At a minimum, this looks like a successful recon action that managed to traverse ~2km of marshland between the river bank and previously held Russian positions. Probably not much more than that for now, but it's certainly notable that this occurred along the route of one of the three bridges that previously traversed the river. One might think that would be an exceptionally well defended spot, but that does not appear to be the case.
 

Incidentally, the location on the left bank of the Dniper where this vehicle was blown up...
...is the exact same spot where Ukrainian forces were reportedly spotted recently.
At a minimum, this looks like a successful recon action that managed to traverse ~2km of marshland between the river bank and previously held Russian positions. Probably not much more than that for now, but it's certainly notable that this occurred along the route of one of the three bridges that previously traversed the river. One might think that would be an exceptionally well defended spot, but that does not appear to be the case.
 

Incidentally, the location on the left bank of the Dniper where this vehicle was blown up...
...is the exact same spot where Ukrainian forces were reportedly spotted recently.
At a minimum, this looks like a successful recon action that managed to traverse ~2km of marshland between the river bank and previously held Russian positions. Probably not much more than that for now, but it's certainly notable that this occurred along the route of one of the three bridges that previously traversed the river. One might think that would be an exceptionally well defended spot, but that does not appear to be the case.
 

Heh, was going to be my comment.  12000 effective fighting force is a freakin division.  A Division on the enemy side of the Dnipro with a secure sustainment link is terrifying for the RA defence.  I am not confident we are seeing that level of operation but a boy can hope and pray.

Pretty impressive counter battery work of a sort - track a GRAD to it's base.  Then destroy everything that's there - vehicles, ammo, personnel.