In the midst of political uncertainty and financial chaos resulting from a pandemic, the pentagon has been busily soldiering on, throwing vast amounts of money toward whatever it wants.
The Trump administration and congress/lobbyist community have broken new ground in procurement corruption by simply ignoring the idea that there are budgetary constraints, or even a budget at all. They’re just spending money like the proverbial cocaine-snorting record executives of the 80s. Except they spend more money in an hour than the total cost of all the cocaine that went up record executives’ noses in the entire glam rock era. [not intended to be a factual statement] It’s tempting to throw a bunch of metaphors and aphorisms into a blender on “frappé” and talk about fiddling while Rome burns, or something like that, but l’air du temps just gives me a splitting stress headache.
To recap: the F-35 is finally starting to have a role in the modern battlefield, and according to everyone who has spent a ton of money buying them, they fly and are able to drop ordnance on defenseless insurgents. They might even be effective in larger battles where state of the art air wings duke it out, assuming that conflict of that sort happens and is not preceded by a great big lopsided beat-down. (which it would be) The F-35’s stealth apparently works OK and it’s a credible first-strike weapon, kind of. Meanwhile, there is the F-22, which is a plane that’s so expensive and high-maintenance that supposedly they stopped making them and weren’t going to make any more, but then the money-valve got jammed open and the defense establishment started thinking “Maybe we could make some of those again and sell them for petrodollars and give sell a bunch to Israel because they asked nicely.”* Remember, these aircraft would be sold to regimes that have already found it difficult to maintain a flying air force of F-16s, let alone 4th generation stealth hangar queens. So, the state of play is that the US has a “troubled” program – the most expensive weapons system in human history – coming in on a wing and a prayer, which means it’s time for the air force to secretly begin developing another fighter jet.
Statue of Kojiro and Musashi in Shimonoseki
Meanwhile, the situation on the ground (or, as they say “in the real world”) is increasingly muddied by drone warfare. Drones are now available in a wide range of sizes, stealthiness, loitering ability, and how many knife missiles they can carry. The cat is thoroughly out of the bag in that department, and military technology appears to be heading toward a crossover-point in which manned air superiority is a relic of The Battle of Britain. For example, in the current conflict between Armenia and Azerbaijan, there are tiers of relatively inexpensive drones over the battlefield. Some are “loitering cluster munitions” – a big prosumer drone carrying a brace of hand grenades – and others are stealthy snipers that carry a single anti-tank missile. Tanks, being optimized for tank-versus-tank battle circa WWII, are locked in a horizontal battle-space and have weak armor on the top, and are hardly stealthy. When your expensive tanks take to the field and start dying randomly from an invisible enemy, it’s really demoralizing to all of your forces. [beast]
Kamikaze drones purchased from Israel have been used to devastating effect by Azerbaijan. These small craft also known as loitering munitions are able to surveil targets including tanks, artillery installations or troops before blowing themselves up. Larger Turkish drones are also flying high above the disputed region and launching missile strikes.
Sure, it’d be important to be able to storm over the battlefield in a stealth jet, and directly drop cluster bombs on all the unfortunates on the ground, but we may be at a tipping-point where conventional forces are just targets and are not cost-effective. I don’t buy the argument that “our men are tough and laugh at explosive death raining on them from the sky.”
In an interview with The Daily Beast, Suren Sarumyan, a spokesman for the Artsakh Defense Ministry, claimed that the Republic of Artsakh has been able to shoot down several drones but he accepted that the unmanned aerial assault vehicles were taking a toll.
“Drones do make an impact on the front line, but our soldiers are among the strongest in the world because they stand firm and fight hard,” said Sarumyan, “The secret to that is that our soldiers defend their home, and it is very difficult to defeat them, even with all the world’s drones.”
Sarumyan’s comments sound like they were made by someone who is living in a nice, safe, bunker with a bunch of concrete overhead.
The premise of F-35s is that they would be doing the same thing as the drones are already doing quite competently. It’s not going to take very long for someone to ask “who needs them?”
Also, as I have pointed out before, aircraft are yummy, thin-skinned, expensive targets when they are parked on the ground. [stderr] The more drones there are, and the more portable they are, the easier it becomes for someone to locate aircraft where they are landed, and attack their bases. The dynamics of force protection are so complex I can’t really say how this is all going to work itself out, but my bet is that fighter squadrons will move farther and farther from the forward battle area, to protect themselves, which makes them less useful since they cannot arrive on target in real-time, while a loitering drone is already there. You know what else is expensive, yummy, thin-skinned and full of fuel? Antiaircraft rocket batteries. I haven’t been able to find it, but a friend of mine sent me a (dead) link to a really expensive Russian antiaircraft (allegedly an S-400) battery getting blown up with an antitank missile from one of those Israeli-made loitering drones.
We’re at a technological inflection-point where the idea of a “forward battle area” and “air superiority” may be breaking down, so that’s exactly the time for the air force to build a new stealth fighter. It’s going to have to be super expensive because it’s going to have to be super stealthy to survive the current environment. It’s going to need some technology so advanced that it’s indistinguishable from bullshit.
The whole thing is super-duper secret, naturally. I believe that is the new strategy, going forward, for escaping any kind of oversight or questions about the program’s inevitable cost overruns. A 6th-generation boondoggle has to not only be stealthy, it’s got to be off of congress’ radar. Not because congress will rein in any such program, but because the inevitable pork-fights will result in the program getting parted out, disastrously, as happened to the F-35. The public does not know if it’s Lockheed Martin, Boeing, or ${whoever} is making it. [defone]
The air force is bragging that they went from a design to a flying implementation in under a year. That tells me a couple of things: it’s got to be leaning heavily on modern aerospace design and it’s probably 3D printed or something like that. One intriguing fact I learned recently is that Northrop Grumman acquired Scaled Composites, Burt Rutan’s company, that designed Project Ares. [stderr, the flying AK-47] If anyone could go from nothing to implementation of a next-generation design, it would be those guys. But they don’t have a history of producing the kind of bloated multi-spectrum boondoggles that the air force wants. Neat purpose-built mini-doggles cannot compete in full spectrum dominance of the 21st century boondoggle space.
Scaled Composites recently released photos of its new 401 prototype (they should have called it the ‘404’ since it’s stealthy) which apparently has some sort of mission for the navy. Interestingly, it’s about the same size and has a similar power-plant to a high-end drone; the line between drone and manned aircraft blurs more. [defensenews]
Just remember all of this when you see American politicians talking about how there’s not enough money to help keep massive unemployment from blowing the economy apart, and there’s no money for protective gear, social services, or any of the other things government is supposed to be doing. There’s plenty of money for next-generation boondoggles and both sides of congress unhesitatingly voted the defense department more money than it asked for.
The way a lot of “defense” gear is “sold” to Israel is: congress grants a gift of money to Israel, said money being ear-marked so they can use it to buy American weapons. That way, US defense industries make loads of money as taxpayers’ dollars are shifted from the left pocket to the right pocket. This has been going on for ages and both parties do it because both parties love money.
“A 6th-generation boondoggle has to not only be stealthy, it’s got to be off of congress’ radar”
There’s something ironically wonderful about a stealth aircraft remaining undetected by the people buying it.
I get the mental image of a company rep gesturing wildly at a piece of empty runway and lots of congress members / media all going along with it, a la “The Emperor’s New Clothes”.
On a more serious note, The Azerbaijan / Artsakh / Armenia seems very unpleasant.
I actually disagree with “Sarumyan’s comments sound like they were made by someone who is living in a nice, safe, bunker with a bunch of concrete overhead.”, even though those assessments are usually very true.
It’s just that the WWI genocide goddamm *looms* over Armenian culture in a way that I never really picked up until I spoke to people in the Armenian diaspora, and I can only imagine that this has intensified since the 2016 conflict and now the 2020 war. I think the speaker was talking in that way because things have not gone so well and because he’s expecting that the Azerbaijani forces will murder as well as fight. Who knows what will happen.
You’re spot on about the tanks though – the Armenian forces have cold-war era equipment for the most part and the tanks of that time are most certainly built with horizontal combat in mind.
I also think you’re right about the drones, but there’s a couple things I’d love your opinion on:
1) Anonymously made drones – ie. supplying useful insurgents with plausibly-deniable firepower
2) Mass drone tactics: Dien Bien Phu style attacks on military bases? (ie. quietly carry in drones until you have a lot of them around someone’s foreign military base / embassy, then unleash all at once).
3) Drone AI – you might not be able to make them do good friend-foe recognition, but you could make what are effectively aerial mines that just home on whatever human-size thing moves in an area, even if you just use them as support while you control the better drones.
4) Some unpleasant person combining the three above ideas.
@aquietvoice:
I’ll have to think about what I can say about some of that. I’m one of those unpleasant people.
I assume that a whole bunch of research these days is going into anti-drone drones. Flocks of them would fly around anything you want to protect, looking for any drone they don’t recognize and blasting it if they see one.
Or possibly blasting random seagulls. The point is, I’m guessing the current drone preeminence won’t last for long.
Poor Armenians, though. Turkey has much to answer for, and the EU should hold them to account.
Whatever whoever is building for however much, may I propose it be nicknamed The Flying Trump? All the winning it’ll do – the craft will probably have a brief tenure because everyone will be so tired of it.
“Scaled Composites recently released photos of its new 401 prototype (they should have called it the ‘404’ since it’s stealthy) which apparently has some sort of mission for the navy. ”
They’ll save that for the MK 4 iteration, which will essentially be an empty hangar (the emperor’s new clothes have already been mentioned) and cost 16 times as much, if my military accountancy and combat pricing techniques are correct.
“The way a lot of “defense” gear is “sold” to Israel is: congress grants a gift of money to Israel, said money being ear-marked so they can use it to buy American weapons. That way, US defense industries make loads of money as taxpayers’ dollars are shifted from the left pocket to the right pocket. This has been going on for ages and both parties do it because both parties love money.”
Tax-laundering.
—
Re: brucegee1962
How would defensive drones address the stealth problem of the offensive drones? For example, it sounds like the anti-tank drones are mainly a problem because you don’t know they were there until you start losing tanks. If you still have to wait for casuaties before you can respond, you could equally well stick to point-defence type systems that are stationary and/or fitted straight to whatever you’re trying to keep safe. Presumably systems like that would be even simpler and cheaper than drones outfitted for the same purpose while also having to fly, navigate and have limited endurance.
I’m asuming that “defensive” implies limited range in the sense that you don’t want to stray too far from your protectee, meaning that mobility gives you little advantage beside some flexibility. The preemptive-strike school of defensive thought might differ.
How would defensive drones address the stealth problem of the offensive drones?
Optical recognition.
You can already buy drones that will image-recognize you and track you. That will get better as algorithms ablnd processor speeds improve.
A future autonomous bomber drone would use GPS to get near its destination then loiter until it sees an F-22 in motion with a heat plume at which point it stoops and drops a load of chopped carbon fiber chaff and ball bearings as it tries to fly into the canopy.
Interceptor drones would optically recognize hostile drones and go all Blendo(tm) on them. Or they might just be spyeyes that recognize hostile drones and call a CIWS (via bluetooth, natch) to hose the area with flying lead.
Re: drones vs. crewed aircraft, it seems to me that even if an aircraft with a human pilot is still much more effective than a single drone, one could buy dozens of drones for the money of one crewed aircraft, so drones are much more cost effective, in addition to not risking a human pilot’s life.
You know, the Pentagon loves having these big wargames where they have a bunch of planes and tanks and people running around in Arizona or someplace pretending to shoot each other. I wonder if they’ve done one of those yet where the red team gets to use lots and lots of drones? Or are we still busily planning on how to win the last war instead of the next one?
Perhaps I’m being too kind, but it sounds more to me like a pep-talk to his own soldiers. “I know it’s tough, but remember that you’re fighting for your homes.”
Besides, what else is he going to say? “The drones are too much to handle. If they send just one more, we’ll all surrender”? Even a standard politician ought to at least know to make the right noises at such a time.
Israel is selling weapons to majority-Muslim countries? I wonder how long that has been going on.
My bet is that the scaled composites thing is a forward drone controller.
Think about it: if you can loiter 100 miles from the action you can have a guy in back commanding drones with low lag-time. Then you need semi-autonomous drones and a framework where the drone pilot can see the battlefield and all the drones on station. Jump into one, fly it for a strike, tell it to go home for reload and jump into another. As the drones’ AIs get better the droneswarm becomes non-linearly more dangerous.
Note: Scott Westerfeld basically describes this in The Risen Empire. If I recall correctly.
Now I’m wondering how expensive it would be to set up a truck with an operator/pilot or two, a decent computer setup, and scores, if not hundreds of practically disposable slightly modified commercial drones, equipped with an explosive charge or similar.
Depending on the area, they may be able to sit back from the actual conflict, fly a drone in close to the ground or well above the fray, and then spiral or zig-zag or whatever it into a bunker, a tank tread, a machinegun emplacement, an antanne/satellite dish, etc. Once it hits its target, or is downed, click over to the next one in the queue, and repeat.
I imagine it could be done at a similar cost, if not less, compared to standard infantry rockets and such. wouldn’t pack quite the punch, but the maneuverability and ability to pilot it from further back might compensate somewhat.
lochaber:
Back in the day, I played V&V (Villains and Vigilantes). One of the rules was the threshold value for damage.
(Or: a hundred fist-pounds against an oaken door will do nothing, yet one bang equivalent to a hundred fist-pounds will shatter it)
Exactly. I could tap you on the chin one hundred times, or punch you there with the equivalent of one taps. Different outcomes.
[one hundred]
Maybe this Slaughterbots video from 2018 really is prophetic: https://www.youtube.com/watch?v=6Ipkq-BASaM
If you haven’t seen it yet, it is quite chilling.
lochaber@#11:
Now I’m wondering how expensive it would be to set up a truck with an operator/pilot or two, a decent computer setup, and scores, if not hundreds of practically disposable slightly modified commercial drones, equipped with an explosive charge or similar.
Let’s imagine a two-tiered structure. There’s a forward controller, and a rear controller. The forward controller is software running on a server. The rear controller is a human strategist that designates and coordinates targeting. The forward controller collects image data from a number of prosumer drones (that have been modified to use decent crypto and to ignore the default vendor-provided software’s GPS and flight restrictions) The drones aren’t “smart” they have some limited autonomy such as flight routing to waypoints, and simple image-recognition for targeting. The forward controller does a lot of image-processing and AI magic fairy dust that recognizes and classifies things, then acts as a sort of cache, building a progressively refined model of the battlefield area. The forward controller has some AI magic fairy dust that allows it to take over one or more of the local drones and bring them under the control of AIs capable of simple mission specifications, e.g.: “go blow up against that fuel truck” or “fill in my view of this map region” – all of this is fed back, post-analysis, to the rear controller’s console, which presents a detailed tactical map of the area with all of the interesting stuff filled in. The rear controller could request image-maps of a region if they wanted to do closer analysis and target designation or tweak the mission-mode of the forward control system or even an individual drone. The rear controller can make various detail-levels/zones of the map available via “push” to other battlefield entities including infantry, artillery, and the loitering (I think they can “loiter” for something like 5 minutes) F-35 strike aircraft.
So, basically, I just described a magical battlefield view/intelligence system, except drone-centric. The current electronic battlefield fantasy that the US is trying to build is, basically, that except it’s centered around other assets like satellite intel and recon aircraft like JSTARS. What I think the US system is not clearly identified as doing is levels of filtering; in the kind of system I’m talking about, you’d have designed-in mechanisms to prevent information-saturation between the forward controller and rear controller. The rear controller would be looking at a screen and get an alert that there was a solid ID on these 4 things up here being tanks and accompanying infantry flagged as hostile, and the human controlling it could sweep/click to designate them as a fire mission for the artillery and air assets.
I’ve played a lot of World of Warcraft and one of the things that I find endlessly fascinating is how our individual brains can make sense of what is going on around us in the middle of an intense raid – but if you watch and you’re not participating, your brain quickly crashes into information overload. [Or maybe that’s just my brain?] I can’t watch the PVP arena battles, and figure out what is going on in real-time but I’ve played a few (I suck) and somehow I can track it when it’s me. All of this is to say that I am deeply worried that the US’ digital battlefield software is not well-designed for tiered information-management. That makes sense, because military commanders tend to be control freaks – if you recall the scene in Black Hawk Down where the commanders in Florida were trying to micro-manage a gun-battle in Mogadishu via satellite link, you know what I’m talking about. I believe that such systems need to be constructed to facilitate common action/reaction loops with a human in the middle. I.e.: you might get a pop up “4 enemy tanks and accompanying infantry identified, 100% confidence, Do the usual thing? Y/N” and you hit “Y” and whatever the usual thing is, it happens. The usual thing might be looking up a table of available assets and posting the information as an action request, while updating everyone else. Back in the day I used to call this “artificial ignorance” for security; a system might observe your actions the first few times you encounter something and eventually start asking “hey I noticed that you like to drop artillery on infantry, from now on should I always queue artillery on infantry if I have a 100% confidence they are hostile?Y/N” When we first started building internet firewalls we used to require the user to sit down and design their security policy, which turned out not to work very well since most organizations didn’t know what they actually wanted. Then a company called Zone Labs came along with a little doo-dad firewall that instantiated that artificial ignorance loop: it didn’t know what a security policy was, either, but it just asked you questions until you built one. That kind of feedback loop is basically what next gen AIs do, they just do it with a lot more complexity, cost, and slower, too. Anyhow, call it “battlefield tactical assist” or something like that. Why not? It’s already happening all over the place: your drones’ flight controller is managing to bank the drone perfectly so that the camera gimbal stays nailed on the target, etc. More software assist and filtering means less battlefield overload and less clusterfuck.
It’s going to be tempting to put the “smarts” at the edge, in the drone. But that will result in very expensive drones – like the stuff the US is building today. I think that tiered controls and accepting that there’s going to be some hard programming to build that infrastructure is the way to go. Then the drones remain cheap and disposable.
In the scenario I envision, a suite of drones might go out and map a battle-area, push the maps back to the controller, and then return for new batteries. Meanwhile, the rear controller marks targets and designs a strike, then clicks the “kill it” button and a second category of bomb-carrying suicide drones start doing low-level attack runs, while a distant artillery battery fires a few 150mm rounds at the harder targets, the whole mess timed to arrive at the same instant.
There isn’t much need for manned aircraft in any of this scenario.
Yeah, but Marcus, you’re talking about achieving surgical victory on an actual battlefield against weapons-bearing opponents. If your strategic goal is blow up whole neighborhoods and city sectors and then act shocked and surprised that you cratered a few hospitals, all while helping your buddies at Lockheed buy their third vacation homes, I assume the F-35 and F-22 will still be your go-to weapons of choice.
The Armenian Azerbaijan
As I understand it the Azerbaijanian offensive, planned to be largely complete in five days, had some initial success and drone did indeed play a part. But after the first few days of getting pounded the Armenian forces retreated to the mountains and have proved to be extremely difficult to shift. As the ground pounders say: ‘it’s hard to break dirt’. And dirt doesn’t get much bigger than mountains.
I wonder about drones. Seems to me that until they have some sort of AI the weak point will be their communications link. Way back we made Tesla coils that could produce powerful and disruptive square waves across a wide swath of the EM spectrum. We shut down after hearing about the FCC seeking out the malefactors disrupting commercial and military communications. Several hundred people were livid at missing several critical plays in a football game and at least one air-traffic-control tower diverted some flights. It made the papers. Rumor was that defective power transformer caused the problem. The POCO denied this assertion … but then again, they always do.
Look for ECM and ECCM, long part of the fight in the air, to become a much larger part of the ground war.
Back in around 2004 I was flying from Seattle to Baltimore and wound up sitting next to a guy who was very military. (If you know the type, you know the type) I introduced myself and we got to talking and he turned out to be the program manager – Lt Colonel Forrest Lindsey – for an experimental artillery system that the USMC was building, called Dragon Fire II. His idea was to use self-loading mortars in light armored vehicles (LAV) that all talked together on a mesh network so they could coordinate fires and even fire on the move to defeat counter-battery fire. This was clever stuff and I was impressed. I was also impressed that he said he “just called Fort Hood and asked the guy who owned the artillery park ‘can I have that Russian-made 120mm recoilless mortar and a few thousand rounds?'” and wasn’t laughed at.
https://en.wikipedia.org/wiki/Dragon_Fire_(mortar)
The part about the system that seemed visionary, to me, was the use of self-organizing systems. Each gun was responsible for notifying all the others as to its current status and location, and there was a layer of “figure shit out” that took place between the guns. He was saying that they had some problems figuring out which gun was in charge and I suggested they use VAX cluster-style hashed elections, and he offered me a job on the spot (which I declined). We had a happy chat about whether or not a CIWS could be used to intercept incoming counter-battery fire – apparently those things are fast enough to hit a 120mm shell inbound (!) but the software linkages they could do aren’t good enough; most military systems are not designed with “composability” in mind. I remember telling Lindsay about the UNIX concept of software composability, and that everything that accepted input should produce output suitable for something that accepted input, etc. It was a really interesting conversation.
So I just looked on the internet for stuff about Dragon Fire II and it sounds like they got quite a ways with it. Here’s a presentation which, I think, may be the one Lindsey showed me on the plane:
https://ndiastorage.blob.core.usgovcloudapi.net/ndia/2004/armaments/DayII/SessionII/07Lindsey_Dragon_Fire_II.pdf
He particularly liked the “drag and drop” interface concept. And I told him the old Robert T. Morris Sr story about early in Gulf War I, and he said that they might have to build that into the interface. So, the story there: I was at a private event in 1996, when RTM Sr was still chief scientist at NSA but the dementia was starting to hit him. Morris told some crazy stories – probably things he should not have talked about – and one of them was about early in Gulf War I, they did not have good enough location data for certain things in Iraq. They had satellite data but they wanted to be able to pinpoint locations with ridiculous accuracy. So NSA whipped up a little box that was innocuous and had a little button on it. They gave a bunch of them to CIA to distribute to assets in Baghdad and told them “go to such and such location and when you are in exactly the right place, just push the button on the little box and keep walking.” One of the assets allegedly asked “what does it do?” and was given the response that NSA had prepared: “God will not love that spot.”
lorn@#17:
Seems to me that until they have some sort of AI the weak point will be their communications link.
Yeah, that’s definitely true.
It’s already starting to work in that regard, though. I flew my DJI drone out of range, once, and it immediately began to return home. The “return home” algorithm is pretty decent, too: it popped up to above normal tree height and then began horizontal beeline to its launch-point. I’m sure that’s all parameterized inside the drone’s onboard controller, somewhere.
I don’t recall if I’ve written about it before or not but the US has a problem with crypto that makes it hard to put commercial crypto in devices – that’s why DoD predator drones use clear comms and CIA predators use commercial crypto: NSA controls DoD crypto applications and DoD would rather have “no crypto” than let NSA get its nose under their tent, etc. Battlefield comms are a fucking bodge of hodge because of cross-domain pissing between rival agencies. I could go on about this at length but really, it’s just a bitter mess.
brucegee1962@#16:
If your strategic goal is blow up whole neighborhoods and city sectors and then act shocked and surprised that you cratered a few hospitals, all while helping your buddies at Lockheed buy their third vacation homes, I assume the F-35 and F-22 will still be your go-to weapons of choice.
Ha, no!
That’s what B-52s are for!
Checkmate, you F-35 proponent!
I’m guessing B-52s fall short in the key strategic goal — the bit I mentioned about the third vacation home for Lockheed executives.
@18 Marcus Ranum,
I assume the guy on the plane did a thorough background check on you and everyone sitting around you before cheerfully spilling the details of our defensive weapons systems?
brucegee1962:
OK you win.
Did you know there is one B52 crew that is the 3rd generation? “Granpaw flew B-52s too!”
I’m surprised the air force hasn’t rebooted the product. “Hey Boeing?!” Maybe repaint some 737Max
I am reminded of a video posted early in the drone age. Someone was flying a drone in a landscape that looked like Southwestern USA. The drone circled a tall rock pillar. When it got around the other side of the pillar, it was out of radio contact, so it engaged it’s return algorithm – which was a beeline to the launch spot. IOW, it flew straight into the rock pillar at high velocity.
I flew my DJI drone out of range, once, and it immediately began to return home. The “return home” algorithm is pretty decent, too: it popped up to above normal tree height and then began horizontal beeline to its launch-point. I’m sure that’s all parameterized inside the drone’s onboard controller, somewhere.
Story I read somewhere, but can’t find now because SpaceX has polluted the sky with shiny satellites the web with references to “drone” and “ocean”: guy takes expensive drone to the beach. Powers it on, flies it over water for a few minutes, brings it back. As he’s about to land it, the battery in his controller dies (or he presses the “return home” button, don’t remember which). Drone shoots up, flies way out over water, comes down, dunks into the ocean… Customer support eventually figured it out: after a cold start with no recent ephemeris data, the GPS can take more than a minute to get a fix. In his case, when it got the fix, the drone was already way out over water, so that’s what it registered as the return-home point. Result: the guy got a new drone, and the firmware was changed to prevent takeoff without GPS fix (unless explicitly overridden; in any case, the return-home function is disabled if the drone takes off before it gets a GPS fix).
Re: #5
Yeees, but that doesn’t really address the question of why you’d need a drone (swarm) in place of some other sytem, even if it’s an old-school fixed-mount, bolt-on, or heavy-thing-strapped-to-unhappy-soldier thingie. Unless you want your “defensive” drones prowling far ahead of anything in need of protection, I’m not sure what exactly you’d gain.
Either way you’re struggling with detection of enemy drones. Something mobile like drones might also end up in the wrong place (e.g. ahead of your asset, where the attacker wasn’t). Drones would periodically run out of fuel or power, while more traditional systems you simply bring along can stay with said asset.
If you’re relying on a vanguard drone defence the area you have to guard becomes increasingly larger depending on enemy range. As I’ve said earlier, you then quickly move into the realm of “pre-emptive strikes” which isn’t defence but obviating the need for defence. Not the same thing, especially if it later turns out it didn’t work.
Incidentally, I don’t share your faith in optical tracking. Building a reliable optical tracking system that works in a combat situation where you may not even be sure what you’ll need to track sounds rather challenging. Given what I’ve seen in machine learning where people actually tried to unravel what was going on, I’m comfortable saying that it’s not the kind of autonomy we want to use right now.* It’s certainly not something we want to connect with any sort of trigger mechanism.
But this did give me a charming vision of a run-away defender lost in a bird sanctuary. It’s not as bad as it sound, though: The laws and rules the US and other nations have in place that allow them to declare civilian casualties to be terrorists after the fact would likely apply to anything avian, too. So, in the end, whatever happened would have been entirely justified.
*Worst** case scenario: SKY as the key identifier for TARGET. Best outcome: Drones race away to destroy the sun, armistice with both sides working together to retool and patch everything so they can go back to fighting each other. Nothing like a common crisis to make peace. Worst outcome: Propaganda war over whose drones got closer.
**As in best.
—
Re: cvoinescu (#25):
That reminds me of the golems of the discworld, who rebelled against their owners by working relentlessly while following instructions to the letter. Nothing does absolute obedience quite like a computer on a mission. It can be maddening at times.
I wonder. If you’re already assuming that counter-drones are possible, possibly using optical systems to bypass stealth, what is to stop you from simply adding these detection systems to good old-fashioned light AA?
I got that reference!
mvdwege@#27:
what is to stop you from simply adding these detection systems to good old-fashioned light AA?
I expect sensor webs that feed back to AA might work, but AA tends to get suppressed by attacks from out of sight. There’s a technique called “fling bombing” that was perfected for that purpose: you fly at the target and before you can even see it, you nose the plane up sharply, and release the bomb on a ballistic trajectory – then turn hard and get out of there. Large bombs can be “flung” for miles.