[createdapril24]

I think the article overstates its premise.

The article makes an assumption that Ukraine has out-innovated Russia in terms of drones. That used to be true (in 2022-2023), with Ukraine's adapting low cost high availability commercial drones as weapons. However since the war has entered the long term Russia's industrial sector has outsupplied Ukraine in terms of quantity of drones, and has reached parity in terms of quality. The realities of the battlefield has seen the cheap and available commercial drones ineffective against mitigations (electronic warfare, surveillance and counter-battery, out-ranging and out-timing). Therefore the contest around drones has moved from quantity of commercial drones to cost-effective quality of drones in the face of countermeasures. This has even pushed some areas of the front to stop using FPV drones (on both sides) in favor of more traditional military drones.

The point of this is actually one of cost. Drones are effective weapons in large part due to their cost-effectiveness. As frequency-hopping modems, larger processors, and multi-frequency antennas are added to cheap drones they start getting expensive - to the point the cost-effectiveness suffers and their drawbacks start becoming more serious.

The article lists drone missions: mine clearing, evacuation, aerial drones, land drones, demining drones, ... . These are mainly overstretches. In terms of how drones have actually been employed in the war, there are naval drones which have been effective at holding Russia's Black Sea Fleet at risk, there are FPV drones which have been used primarily to stall the frontline defensively, and there are long range drones (not mentioned) that are used as an alternative to e.g. ballistic and cruise missiles.

There are R&D projects, many of them failed at employing drones for many missions. There are some partial successes in using them to lay mines (and fake mines). But there isn't as much success in using them for de-mining. Lifting a person for evac is very challenging to do with a drone, especially an autonomous one. Somehow "surveillance" doesn't make the list, but probably the #1 contribution/mission of drones is battlefield visibility.

The article compares taking 200 artillery shells to kill a building, vs 1 drone to kill a soldier. This is not only inaccurate, but it's apples-to-oranges.

Drones have not replaced artillery in the Russo-Ukraine war and they aren't going to. Drones cannot be massed because they interfere with one another in the electromagnetic spectrum. They also require a significant number of people to operate compared to artillery. The munitions on drones are far less powerful than artillery (e.g. drones have trouble destroying armor, artillery doesn't). Simple means (mesh, nets, smoke) can shut down drones - but have nothing on artillery. There's so much to say here but it's not even comparable and the article bases its primary takeaways from this incorrect assumption.

The article discusses that drones can fly "up to 22km". It doesn't mention that these require re-transmitter drones, which need separate pilots, separate modems, logistics coordination and ultimately - a much higher price tag. It can be worth it for certain missions, but it's hardly true that innovation has somehow created drones that are just better units. It's more that the employment of drones in warfare has gotten better - the command and control has gotten better.

The artificial intelligence aspect of drones is typically a "terminal flight system", one that can take over to seek non-moving targets once EW has shut down communications. While its true that Russia and Ukraine both use terminal flight guidance systems to deal with the "last mile" of EW cover, its nearly impossible to use this to hit moving targets or to hit vehicles with armor in the areas that are needed to achieve disablement or a kill.

The DoD has applied Project Maven to solve the problem the article discusses: automatically identifying targets to strike (which honestly applies to more than just drones - can be used for artillery, etc). Unfortunately Project Maven has been disappointingly significantly less accurate than human analysts at identifying battlefield targets on the same imagery.

I could go on, but I think the article is kind of stuck in 2022?

There's a certain pitch one can make for drones, and Ukraine is making that pitch. But I suspect the author might have too narrow a set of sources or some kind of biased interest, for what they are writing about.

[btbuildem]

> Drones cannot be massed because they interfere with one another in the electromagnetic spectrum. They also require a significant number of people to operate compared to artillery.

I think we've all seen these drone shows, where a large swarm maneuvers to make shapes in the sky, right? We're not far at all from one operator controlling a swarm, capable of pre-programmed formations and maneuvers.

The economies of scale work overwhelmingly in favour of drones -- small, light, disposable.

[createdapril24]

In theory, sure, but those swarm shows are pre-programmed: they aren't reacting dynamically to a (chaotic) situation on the ground and they aren't communicating with operators.

There's some far-flung future in which drones are fully autonomous and in fact don't even need antennas. At that point it's possible they can be massed. But it's a bit of a science fiction. At least, there isn't any such product available (commercial or military) and this isn't how drones are being used in war.

For economies of scale - true also of artillery and other equipment! The more you can scale production the more cost-effective the weapon. As mentioned, drones in the Russo-Ukraine war are starting to see their cost-effectiveness wane due to having to become larger (larger munitions, large antennas, etc), heaver (bigger batteries, larger munitions, etc), and non-disposable (high cost frequency hopping gear, difficult to find munitions, difficult to source batteries, etc).

Giant technological leaps could happen, but it's nerve wracking to bet the outcome of a war on something like that.

[yencabulator]

Meanwhile, the front line troops are reporting that almost all the drones are FPV guided. Computer vision apparently just isn't good enough to fly in through a doorway, like the FPV drones are used.

(Sure, you could make a cluster of GPUs do that, in a test. It's completely different to do it in war.)

[ak217]

I suggest following the thousands of drone videos that continue to come out of the conflict. They contradict much of this analysis.

Small and medium drones are much more precise than artillery and remain incredibly effective at surveillance, spotting and at attacking the poorly operated, low survivability armor that Russians have, despite countermeasures. Ukraine is now domestically mass-producing bomber UAVs able to deliver multiple mortar rounds semi-autonomously on a target. While ECMs (jammers) have undoubtedly reduced drone effectiveness, ECMs come at a cost. A small ECM will have a very localized effect, while a large ECM can be taken out with anti-radiation missiles, as the Ukrainians have been doing.

Everyone is also recognizing the huge psychological effect of drones in the battlefield now. Any kind of massing is getting increasingly risky, communications, autonomy and mobility become critical.

The main point of the article is that the war is driving innovation in drones, which is pretty self-evident and doesn't seem overstated at all.

[createdapril24]

It needs to be noted that these drone videos are highly filtered - they are used as fundraising tools by Ukraine Armed Force units and then what "trends" are the spectacular videos (and often are out of context).

So while you may see videos of drones taking out tanks, they are often tanks that have already been killed and drones come in on an immobile target, etc.

You would also think that Ukraine is using more drones than Russia - based on the videos. But Russian Armed Forces don't need to fundraise for their equipment on Telegram, and what videos of Russian drone attacks that are published don't make it to trending in American social media (because "disinformation").

What's happening on the battlefield and what you see in those drone videos are very different due to selection bias.

Drones have less of a psychological effect than, say, glide bombs or artillery. This is due to the size of the munitions.

The broad thesis that war is driving innovation in drones isn't incorrect. But the supporting material in the article is (I've listed some of them above, and you seem to agree with them, point by point?). The article is overstating the case for drones, even calling them the successors of artillery. That's all my comment is addressing - let's right size and calibrate this: drones are being innovated on, however they are less effective at higher expense than they used to be, they very well may continue to trend in that direction, and they aren't a replacement for conventional weapons.

I guess another way to put it is the title of the article isn't wrong. But if you read the article, the content clearly is.

[pigpang]

> So while you may see videos of drones taking out tanks, they are often tanks that have already been killed and drones come in on an immobile target, etc.

Those tanks are immobilized by other drones or mines.

> You would also think that Ukraine is using more drones than Russia - based on the videos.

Yep, Ukraine uses more drones than RF in both absolute and relative terms, because Ukraine uses 10x less artillery shells.

[ak217]

No, I don't really agree with any of your points, I think they're either incorrect or not relevant. And the rest of the content in the article beyond the headline is reporting, not making a value judgment, so I don't see how it's wrong.

[createdapril24]

Oh interesting. No rebut of specific points or facts, just blanket disagreement?

I think its hard to call a whole article "correct" or "incorrect." This article makes many assumptions, states many things as facts, and yes - makes judgements. For the reasons states above I think the article overstates its case. It's behind the newest reporting by around two years. It misses or completely excludes important details (e.g. range requirements for retransmitter drones).

That doesn't make the article wholey incorrect or worthless. But yeah I think if someone is reading that article they should be aware of its inaccuracy, factual deficits, etc...

Happy to dive into any of the one-by-one points in the GP comment if there's questions or skepticism.