kimbosbread

Oh yeah, I was thinking in general, not the Kerch bridge. How does one degrade 100-1000s of km of railroad without spending ****loads of money, and with minimal risk. The point of it is that Russia without rail will collapse. A siege of Crimea might take a while. If you can degrade Russian rail and thus society, military, push ethnic republics to "alternate" arrangements a la Galeev, that's a big win. 100%

There is nothing I know that exists currently that does this out of the box. To degrade the system without the pain of destroying the rail: If you knew where all the signals and control boxes were (gps) that are difficult to replace, and had suitably long range small drones with 1kg HE, you could cause significant degradation to the rail network, but that's because you are knocking out difficult to replace stuff all over. Alternatively, hit the rail yards and destroy enough locomotives, again with your long range drones/cruise missiles/saboteurs. Let's say we want to take out a 100km line of rail. That's a ton of steel, wood, dirt and gravel to move, and it's easy and fast to fix.I can imagine two possibilities to "move": [Mechanical] A track unlayer, which rides the rails and somehow tears up the track behind it (explosive, or just mechanical) [Explosive] Drones that drop charges at periodic intervals along the track (say every 5-10m), for several km. Agricultural drones already can drop/plant seeds in complex patterns; in this case it's a matter of can you carry enough charges that each can cause serious damage to a section of track. Imagine an airborne UR-77, if you will. I don't know enough about trains to know if there's some other target; I guess if you had time you could wait for them to run out of spare parts.

What we need is time. They have about 10 years of demographic runway left. They know, we know and they know we know. We need both credible deterrence and the capability to fight a short war and a long war. China is betting they can win a long war. All the missiles get used, and the ships get sunk? They can build more, fast. Us, not so much. China's assumption is that they can win if they can get troops to Taiwan and hold for two weeks. My worries: Does the US have a big enough submarine force? We have 53 fast attack boats; lets say emergency we have half those are in the area. Do they carry enough weapons to sink a ****load of Chinese ships? Los Angeles and Virginia class carry 26 torpedoes (and up 12 missiles for some boats). Seawolf class carry double that. That's 1300+ weapons. Chinese navy is 300+ armed vessels and growing fast; plus they have lots of merchant ships. We can sink of bunch of ships, but they'll have more. How do we resupply our subs in a timely manner? Given Chinese antiship missiles, can we get surface vessels anywhere close to Taiwan? Are large surface ships on a blockade near Singapore viable? Would we be better off playing pirates on RHIBs in those sorts of waters and moving the surface ships back to the Indian Ocean/Australia/Indonesia? What do we do about drone ships/subs/mines? We need to be building these, fast, but I bet China can do it faster, and better. If we need to boots on the ground in Taiwan, can we get marines 400 miles from Okinawa in a contested environment? What do we do about transiting the Panama canal, which China effectively owns nowadays? On the other hand: If all the satellites get taken out day 1, we have a credible shot at getting a replacement constellation back up, fast. China does not have a counter to this besides building more anti-satellite missiles. The longer they wait, the more ridiculous our space overmatch becomes. China imports food + fuel. How much can they stockpile? That dam is a big target and single point of failure. They'd lose Wuhan, Shanghai, and bunch of power plants, tons of military bases etc. Are they willing to risk it? Also, I used to work with a bunch of mainlanders who are all very pro-China. Around 2021 they had kind of given up on Pooh-bear and were much more bearish, no pun intended.

If you are any good, why the hell would you join the military to be an IT guy? If you wanna be a soldier sure, but IT... what's the point? If you are even halfway decent, there are thousands of tech companies, many in military adjacent areas that pay well enough to afford an apartment in a tier1 US or Candian city without roomates (or in SpaceX's case, half of a FAANG with longer hours). You don't have to be a stanford grad or anything like that either; it's purely ability to pass coding and design interviews at the entry level, which I've seen average but motivated people do easily.

In terms of detecting potential leaks, I would hope the approach is similar to what we do in the cybersecurity industry: Essentially have model of normal user or process behavior, and then for every behavior or non-behavior that falls out of that scope, we increase the risk score.

Yeah do we have a giant moving castle, or should we be a submarine? Or do we go the 3rd way and go full Zerg, since if you can't beat them join them. Beyond the cost of adding $500-$Nmillions of APS for each truck, the ammo is such a big problem. The economics just suck if the APS is more expensive than a bunch of munitions. An FPV drone with a bunch of explosives on it is really cheap, and really accurate. And this isn't even touching on munitions that do their thing outside of APS range, or swarms combined with that.

Re Trophy, I don't think that's a good long term solution, especially against smarter munitions that don't just attack from the sides, let alone a grenade dropped from a drone. It's definitely not gonna handle multiple drones/missiles attacking simultaneously from various angles.

Unrelated to the mole hunt, I was wondering about the application of drones to engineering tasks, namely mine clearing. I know there has been a ton of work planting trees with drones (Airseed, Flash Forest, etc.), so I wonder if you could carry enough mine-setting-off-things on a big hexcopter (payload of 10+kg) and have it drop these at regular intervals on the suspected mine field you plot out on GPS.

If I were China, I would set up WallStreetBets, but for classified info. Instead of losing money/betting big porn, it's leak the most classified/ridiculous info porn and improve the size of your e-penis.

The great thing about the plan is even partial success directly threatens the railroad to Crimea, putting huge pressure on Russia. Obviously there's a question of cost, but geopolitically besieging Crimea seems like a big deal to me. It is huge for the domestic audience, puts enormous pressure on Putin, and isolates a major portion of Russia's forces. Best case: I think the most successful outcome- pushing Russia back to the 2014 Crimea border, dropping the bridge, and taking back Melitopol is a huge win. It's a ton of territory and might well lead to a Russian military or political collapse. Crimea is the only part of Ukraine Russia wanted in 2014 enough to formally annex it. Medium case: Let's say Melitpol is not retaken, and only the western side of Kherson. That still doesn't increase Ukranian frontage much, and puts tons of pressure on Crimea. Worst case: Either Vasylivka or Tokmak or Polohy are the only places retaken, that barely increases frontage for Ukraine and still threatens Russia's rail line to Crimea. And then the pivot to Luhansk north is easy. Suppose Ukraine is able to barge/bridge at Kherson rapidly and build a 10km bridgehead. Does Russia have any weapons that could threaten a crossing once established, especially if there's distraction at Tokmak?

I very much like the idea of Troitske and hugging the river on the southern front and going for Tokmak + Vasylivka instead of Melitopol. Strongly agreed. Ukraine needs to avoid getting bogged down and instead capture territory and hammer logistics and command, making it unteneable to hold the city. Also there is the moral effect- Ukraine reconquers a city, it is not completely destroyed, vs Russia it's complete destruction. Is there some other major goal that would justify an attack nearer to the "center" of the front? The north and south attacks basically serve to cut railways (one that isolates Crimea, the other that makes the entire SMO harder to resupply). I cannot imagine any reason to go for Berdiansk or Mariuopol over going West; similarly Donetsk is presumably too fortified to be worth it. The only thing I can think of is an opportunistic encirclement or flanking maneuver of exhausted Russian troops that have been basically enjoying the freeze-thaw cycle for the entire winter and early spring. What about going for Lychansk and threatening the rear of the Russian forces at Bakhmut?

It's gotta be a 1-2 punch North and South to force Russia to make a choice. For the former, Starobilsk makes the most sense to me in that pushes Russia's rail suppy lines another 50+km to the east and will worsen the resupply situation that much more, plus directly threaten Luhansk. How is the dam at Nova Kakhova nowadays? Would that be a viable surpise crossing point?

It's also pushing us towards ChatGPKill, which was gonna happen anyway, but will have to happen faster if we lose all our pilots and boat people in a war with China.

Sealift + airlift + logistics personnel + training + ISR assets in air and space + missles like ATACMS. It's a safe assumption that a lot of this equipment would get blown or used very quickly, and any of it in Ukraine would be hindering an effort in the Pacific.

You don't need a gun at all. Just go Phantasm style and put a drill with an electric motor on the front of the drone and fly straight into the target's forehead and drill baby drill. You can cobble together incredibly good face/head tracking software that will run on minimal hardware.

Me too. I doubt we'll have to wait long though. On a similar subject, I wonder if there's gonna be a rapid evolution in sniper rifles soon, where the spotter + sniper converges with kamikaze drones.

The US should show Europe who is boss and admit Ukraine + Moldova + Georgia to NAFTA. Forget the EU, our club is more awesome plus we have poutine and tacos.

I suspect standarization is key. If it's a built in battery at the factory, at least same charging interface. If it's a fuel, same fuel. I wonder if otto fuel (from torpedoes) would be useful, especially for single use drones? Yeah if your per drone cost is $1000, but you have a good kill/useful intelligence ratio every time you consume a drone, maybe it's not even a driving consideration. So for example if you decide to deploy your drone swarm in area X, do you consider that to be the same as CAS where let's say you've burned $300k of fuel, maintenance and airframe (or the equivalent in humvee hours and associated back surgery/physio for the crew), so you just say whatever, all the drones are considered consumed? It's really about the choosing which of those conflicting requirements are the most important. Do you care about reuse? For micro cruise missiles, I suspect not. It's basically a flying smart mine that can coordinate with all the other smart flying mines without phoning home. For Mr Robotank, I suspect reuse is much more important, as well as maintability and fitting into the supply/maintenance organization with wheels, ammunition, fuel, poorly applied reactive armor tiles etc.

Rechargeable means only using half of your battery charge every flight, unless you are find junking that battery in 100-200 use cycles. If you have small drones (fixed wing, at least not power hungry quadcopters) with longer ranges they can make their way much further back to find a recharge pad (let's say no battery swap) or for a refuel. Nice thing about sealed rechargeable battery is you could just slap an induction pad on it and although charging is slower, there is no port for dirt to get into. It also means you could just drop a bunch of passive self contained recharge pads on the battlefield that would be good for 100 charges and you consider them a consumable (a nasty, toxic lion one though). Swapping batteries is much faster if you can figure out a good autonomous way to do it.

As a relatively uninformed software engineer, I think this is where things will go. We've been able to build small (10cm long) fixed wing drones that can fly for an hour at 100km per hour for 24+ years with a camera (at least when I was visiting the MIT aerospace labs in 1999). You can build something like this now for very cheap, and much improved performance just on the hardware side. Let's say these are electric drones, so trade endurance for no need to truck fuel around and simpler motors, you could have thousands of these things in the air, with 25g of high explosive in the nose. Combine that with modern ML/image recognition software + GPS running on hardware that fits into a tiny package AND is cheap and easy to build, as in $100 or less cheap. Effectively this is a smart cruise missile scaled down to kill an individual soldier. Let's say Ukraine could have a ten thousand semi autonomous drones roaming 20km behind Russian lines looking for targets for several hours at a time; would you even need mass if it was cheap to dedicate a drone to kill each enemy soldier? If the defences are cheap enough to scale with the number of drones, maybe. Lasers seem reasonable, but they require a lot of power and cooling, especially for longer range. As soon as a laser is identifier, something specialized can be sent after the laser. The more insiduous thing with drones and small computers is there is no reason the drones can't be used to execute flexible tactics. For example, we've identified a laser on a bmp-9000, and we estimate it can only fire 20 times per minute, so let's swarm it with 25 drones at once, or attack from multiple directions at once with 5 drones.