How does todays tech look like? The previous trend was scooping from far above with a mid size drone, but I guess nowadays the insectocopters are possible.
Or did spying on phones make this absolutely redundant :) ?
The PD-100 made by Prox Dynamics is a modern reconnaissance drone which is close to the insectothopter in size. It's pretty cool to watch in real flight, as it is very difficult to notice when hovering just a few meters above you.
The main efficiency boost is using a helicopter style rotor over a quadrocopter - the larger propeller means greater efficiency.
Obviously a great deal of time is spent on weight reduction. And finally the max range is one direction, until the battery is completely dead. When you have military funding there's no need for a return trip... and to think each of these could send a kid to college.
I was super disappointed after having read about this in books/news for years being referenced as in the spy museum, and finally got a chance to visit (it takes living in DC area to fully experience DC) only to realize it wasn't available for viewing.
More on topic, the real discussion this piece of work enables for me is the reduction in the gap of power between the government and the individual citizen due to technology as a force multiplyer. So they used to say the gov was ~10-20 years ahead of the civilians. I think the gov is more like 5 these days. It's corruption has weakened it's effectiveness.
You get $40 toy drones the size of a hand with HD video today. I keep telling people that interesting things will happen when these shrink enough and get enough flight time for people to decide it's low enough risk to start flying drones into restricted areas and buildings.
It'll be increasingly hard to defend against, not least because it's getting cheap enough you're not going to be defending against a handful of enemy states, but against anyone curious enough or with a sufficient grudge and a willingness to take the risk.
The challenge is the toys are getting so cheap, that it is likely to become highly difficult to do this for a committed adversary.
E.g. imagine just overcoming the shotgun or net drones by simply sending in a thousand (or however many) toy drones to deplete your supply of nets first. Of course it'd set off all kind of alarms, but that may in itself make it worth it: How many people can you afford to keep on high alert for $12 [1] disposable drones in case one of them turns out to be something that actually matters.
The potential disruptive effect of just toys is becoming massive.
Then multiply it by a high factor if adversaries in that swarm of toys hide e.g. a handful with explosives, or a handful of "proper" surveillance drones, so that the cost of not taking out every single one (or even of taking one out in the wrong location) becomes prohibitive.
[1] https://www.alibaba.com/product-detail/2-4G-6-Axis-RC-Drone_... - $12 if you order large enough volumes; of course these would not do it - for starters, you'd have no chance to control all of them, and they could be disrupted simply by transmitting enough noise at 2.4GHz. But replace the control of it with some semi-randomized flocking on all of the disposable ones, and cause chaos.
The problem, though, is that a $12 toy drone won't be able to reliably stop even another $12 toy drone, and you need to be able to stop something much bigger and more sophisticated. Because you won't know if what's coming is harmful or a major security threat.
So you might very well look at a $100k drone designed to take out $10k threats being used against a $12 toy because the only way of being sure is to take it out.
Now suddenly an adversary able to send 1000x$12 drones for $12k materials cost could conceivably force you to spend $100m on drones for a single facility.
The problem with that being that your adversary gets to decide where to attack.
That's the nightmare of this: Highly mobile, extremely cheap "attacks" that can be used to massively drive up the needed defense efforts on the off chance that somewhere in a swarm of a 1000 the adversary has hidden a $10k investment in a flying bomb.
A superior budget may win, but this can drive up the budget advantage needed before you have an advantage.
I’m pretty sure a $10,000 drone can hunt down a $12 drone, and do it more than once. To make it work you need size, speed, radar / sonar, a computer, and a weapon for that sort of drone. Civilian DSLR-carrying drones are half-way there, they have the size, the speed, and the weight capacity. Don’t know the size/weight of the sonar though.
Besides cheap drones have low range and low speed, it will be hard for the perp to get away, or even get close enough in the first place.
Non-state actors are very limited in budget. They might catch some targets off-guard, but it’s not a game changer.
There is also a small radar inside the hellfire missile, but I don’t know the weight. I doubt it’s too heavy. And hellfire are $100k a pop, which is why I think radars shouldn’t be too expensive.
This misses the point. A $10k drone can probably take out a $12 drone, but you need to be armed to take out the bigger threat, because you don't know whether the drone any given defensive drone will try to take out is a $12 drone or another $10k drone with defensive capabilities of its own.
That's the point of this scenario: Deploy a bunch of cheap junk, in between a handful actually dangerous ones. If you knew they were all just $12 drones you'd hardly even need a countermeasure - they'd fall down after not that many minutes flight.
The point would be to force the defensive force to commit far more resources than necessary.
> and do it more than once.
Yes, but the problem is you need to scale for peak load. If someone sends 1000 drones after you spread around your perimeter and it'll take those drones 5m to target, you need to have enough to intercept and destroy all 1000 in 5m. Sure, you can reuse many of them - next time your adversary may up the attack to 2000 drones, and see if you still have enough to kill them all.
> Besides cheap drones have low range and low speed, it will be hard for the perp to get away, or even get close enough in the first place.
The $12 drones in question have wifi and transmit hd video. You can get suitable SOC's with dual wifi chipsets and relay connections quite easily, and some of those cheap drones also have proximity sensors good enough to be able to add tolerable flocking behavior. Judging from the stability of automated takeoff and "return" functionality on the toy drones I've tested, we're not long away from having the ability to have them do unaided flight. Not least because the toy ones do not need to reach a specific target: Flying erratically is not a problem. They just need to cause fear. Speed is also less of an issue if your intent is to spread fear - you can fly drones low over locations where a defender does not dare use too drastic interception means for fear of causing more harm than they prevent.
Besides, the most likely use for this type of thing are in situations where some desperate group is fighting a state actor and losses are high anyway. It doesn't make it worse.
> Non-state actors are very limited in budget. They might catch some targets off-guard, but it’s not a game changer.
I think it is, exactly because they are very limited in budget; the terror you'll spread with a $10k explosive device + $10k worth of drones aimed to make the response more expensive is likely to be far greater than what you'll achieve with 2x$10k explosive devices, and the chance of actually hitting anything of value can likely be made higher. E.g. look at the rockets regularly fired by Hamas - it's rare they cause much real harm, because they're basically unguided; they're mostly about fear. For that kind of scenario, the ability to get people to fear the buzzing of a drone because they don't know if it's a $12 toy or a $10k bomb would definitively be a game changer.
You can call this pure conjecture if you will, but here's the story. Several years ago, I knew some people in a PhD physics program, and one of the guys there tells a story about an engineer he worked with during one of his summer internships during undergrad. He described this guy as one of those older rugged not quite polished types. The type that cursed a lot, and didn't really put up with anyone's crap. He was wicked smart though, and he'd been doing it a long time too. The story goes that the one of the government agencies approached him to help them work whatever they were doing at the time. I think he said this was around the 1950s. The engineer asked why should he work with the government when he gets to work on cutting edge stuff all the time anyway. Apparently, they let him see some of the things they had developed to entice him, and he said that it was easily 5 years beyond what anyone had seen.
Is it true? Honestly, I don't know, but the guy telling the story at the time believed him, and this was in the same conversation where scientists were talking about how they could use satellite imaging to track where someone had walked 10 minutes before in a field. That was the unclassified version of what they could talk about.
Has any of this tech "trickled-down" to agricultural applications? The ability to track an individual seems like it could be used very effectively for livestock monitoring.
It's probably not as effective as an ear tag and fence though (especially when you factor in real world conditions, a lack of unified technologies across any industry, and an unwillingness to invest in technology).
That said, I have heard of remote sensing being used to monitor livestock and crops. The commercialization of UAV's is contributing to that.
I remember when astronomers started using lasers for distortion canceling. A buddy of mine had been working on spy sats for a decade and said the government was using the same technology (though in the other direction) when he arrived.
Adaptive optics came from the military this wasn’t a secret.
Hubble was built on a spy satellite bus.
Military gets the shiny toys first the reason why we have such great gyro sensors today in phones is because solid state sensors were first developed for guided missiles.
Is that because MEMS is cheap (TCO, not BOM) and good enough or is it because MEMS is better? It's obviously not as good at drift, but I'm sure there are other dimensions of performance.
Good enough MEMS aren’t good as laser ring gyros, a single sensor isn’t even good as fiber optics but multiple sensors with per sensor drift bias taken into account is good enough.
You can’t strap a 150,000 INS on every shell you can’t even strap a 30,000 FOG INS on a shell.
Basically since we now want INS and GPS on every bloody field equipment and anything above a 20mm grenade should be guided muinitions you have to make some compromises.
And MEMS aren’t that bad yes they drift but for their usecases the drift is negligible.
Good, we're on the same page. Any chance you know what's holding up integration of RLGs? I'd have thought they would be a natural fit for a cheap silicon implementation.
Because RLGs are still spin gyros the dithering motor has to spin and spin pretty darn fast (to prevent lock-in) which prevents miniaturization.
MEMS operate on vibration which can be induced with solid state ultrasonics.
Also since most navigation systems today are hybrid drift isn’t much of an issue over long distances anymore and for short distance applications it doesn’t matter.
MEMS are much smaller, FOGS can also be made smaller especially considering you can use a single external laser injector to feed multiple sensors.
So while RLGS are still probably the most accurate production gyro we can make there are technical issues on making them smaller and since we have alternatives there isn’t likely much reason to invest in miniturizing them further if it’s even possible.
Given the subject of the article and some of the comments (robotic or modified real animals) I was expecting this to be about a different type of seal.
What an interesting life you must have, I thought, if people gift you bugged aquatic mammals as pets...
The technology will continue to improve. At some point they'll attach a hypodermic loaded with cyanide or ricin to one of these and use it in an assassination.
Exactly. It seems the only thing preventing countries from war or assassination is the fear of reprisal. If the US could send 50 of these to take out the top NK military it's hard to believe they wouldn't take that risk.
Range is an issue, so you’d need an agent to smuggle it in country, verify the targets and approach, then release. Now, if they’re caught before, or if the device is recovered... game over. A super high tech assassination speaks to the identity of the assassins, risks leaving evidence, and worse, giving up the tech.
Why? Just make the current iteration have a sharp point on it and coat it with your desired toxin or tracker. The tech is here now, and we’re in real trouble without solid laws around this.
The tech is here now, and we’re in real trouble without solid laws around this.
People who would be interested in the misuse of this technology aren't, as a rule, interested in whatever laws you might think appropriate for regulating it.
What's interesting is the engine for the insectothopter, it appears to be a plastic bladder compressed with a cantilever spring. Inflating the bladder with gas pulls the wing down, which is itself something like a leaf spring, and then deflates and the wing springs back. This is a surprisingly simple mechanism that one might even be able to make without hand tools.
I wonder how long it will take the RC community to duplicate it?
53 comments
[ 5.4 ms ] story [ 58.1 ms ] threadOr did spying on phones make this absolutely redundant :) ?
http://www.proxdynamics.com/home
Obviously a great deal of time is spent on weight reduction. And finally the max range is one direction, until the battery is completely dead. When you have military funding there's no need for a return trip... and to think each of these could send a kid to college.
https://en.wikipedia.org/wiki/Black_Hornet_Nano
More on topic, the real discussion this piece of work enables for me is the reduction in the gap of power between the government and the individual citizen due to technology as a force multiplyer. So they used to say the gov was ~10-20 years ahead of the civilians. I think the gov is more like 5 these days. It's corruption has weakened it's effectiveness.
You get $40 toy drones the size of a hand with HD video today. I keep telling people that interesting things will happen when these shrink enough and get enough flight time for people to decide it's low enough risk to start flying drones into restricted areas and buildings.
It'll be increasingly hard to defend against, not least because it's getting cheap enough you're not going to be defending against a handful of enemy states, but against anyone curious enough or with a sufficient grudge and a willingness to take the risk.
Meanwhile the Police in Holland have abandoned an attempt to use Sea Eagles https://www.google.ch/amp/s/www.theverge.com/platform/amp/20...
E.g. imagine just overcoming the shotgun or net drones by simply sending in a thousand (or however many) toy drones to deplete your supply of nets first. Of course it'd set off all kind of alarms, but that may in itself make it worth it: How many people can you afford to keep on high alert for $12 [1] disposable drones in case one of them turns out to be something that actually matters.
The potential disruptive effect of just toys is becoming massive.
Then multiply it by a high factor if adversaries in that swarm of toys hide e.g. a handful with explosives, or a handful of "proper" surveillance drones, so that the cost of not taking out every single one (or even of taking one out in the wrong location) becomes prohibitive.
[1] https://www.alibaba.com/product-detail/2-4G-6-Axis-RC-Drone_... - $12 if you order large enough volumes; of course these would not do it - for starters, you'd have no chance to control all of them, and they could be disrupted simply by transmitting enough noise at 2.4GHz. But replace the control of it with some semi-randomized flocking on all of the disposable ones, and cause chaos.
So you might very well look at a $100k drone designed to take out $10k threats being used against a $12 toy because the only way of being sure is to take it out.
Now suddenly an adversary able to send 1000x$12 drones for $12k materials cost could conceivably force you to spend $100m on drones for a single facility.
The problem with that being that your adversary gets to decide where to attack.
That's the nightmare of this: Highly mobile, extremely cheap "attacks" that can be used to massively drive up the needed defense efforts on the off chance that somewhere in a swarm of a 1000 the adversary has hidden a $10k investment in a flying bomb.
A superior budget may win, but this can drive up the budget advantage needed before you have an advantage.
I’m pretty sure a $10,000 drone can hunt down a $12 drone, and do it more than once. To make it work you need size, speed, radar / sonar, a computer, and a weapon for that sort of drone. Civilian DSLR-carrying drones are half-way there, they have the size, the speed, and the weight capacity. Don’t know the size/weight of the sonar though.
Besides cheap drones have low range and low speed, it will be hard for the perp to get away, or even get close enough in the first place.
Non-state actors are very limited in budget. They might catch some targets off-guard, but it’s not a game changer.
There is also a small radar inside the hellfire missile, but I don’t know the weight. I doubt it’s too heavy. And hellfire are $100k a pop, which is why I think radars shouldn’t be too expensive.
That's the point of this scenario: Deploy a bunch of cheap junk, in between a handful actually dangerous ones. If you knew they were all just $12 drones you'd hardly even need a countermeasure - they'd fall down after not that many minutes flight.
The point would be to force the defensive force to commit far more resources than necessary.
> and do it more than once.
Yes, but the problem is you need to scale for peak load. If someone sends 1000 drones after you spread around your perimeter and it'll take those drones 5m to target, you need to have enough to intercept and destroy all 1000 in 5m. Sure, you can reuse many of them - next time your adversary may up the attack to 2000 drones, and see if you still have enough to kill them all.
> Besides cheap drones have low range and low speed, it will be hard for the perp to get away, or even get close enough in the first place.
The $12 drones in question have wifi and transmit hd video. You can get suitable SOC's with dual wifi chipsets and relay connections quite easily, and some of those cheap drones also have proximity sensors good enough to be able to add tolerable flocking behavior. Judging from the stability of automated takeoff and "return" functionality on the toy drones I've tested, we're not long away from having the ability to have them do unaided flight. Not least because the toy ones do not need to reach a specific target: Flying erratically is not a problem. They just need to cause fear. Speed is also less of an issue if your intent is to spread fear - you can fly drones low over locations where a defender does not dare use too drastic interception means for fear of causing more harm than they prevent.
Besides, the most likely use for this type of thing are in situations where some desperate group is fighting a state actor and losses are high anyway. It doesn't make it worse.
> Non-state actors are very limited in budget. They might catch some targets off-guard, but it’s not a game changer.
I think it is, exactly because they are very limited in budget; the terror you'll spread with a $10k explosive device + $10k worth of drones aimed to make the response more expensive is likely to be far greater than what you'll achieve with 2x$10k explosive devices, and the chance of actually hitting anything of value can likely be made higher. E.g. look at the rockets regularly fired by Hamas - it's rare they cause much real harm, because they're basically unguided; they're mostly about fear. For that kind of scenario, the ability to get people to fear the buzzing of a drone because they don't know if it's a $12 toy or a $10k bomb would definitively be a game changer.
Is it true? Honestly, I don't know, but the guy telling the story at the time believed him, and this was in the same conversation where scientists were talking about how they could use satellite imaging to track where someone had walked 10 minutes before in a field. That was the unclassified version of what they could talk about.
That said, I have heard of remote sensing being used to monitor livestock and crops. The commercialization of UAV's is contributing to that.
Hubble was built on a spy satellite bus.
Military gets the shiny toys first the reason why we have such great gyro sensors today in phones is because solid state sensors were first developed for guided missiles.
MEMS gyros are used in military INSs today can’t say if in missiles but in guided mortar shells and field artillery for sure.
You can’t strap a 150,000 INS on every shell you can’t even strap a 30,000 FOG INS on a shell.
Basically since we now want INS and GPS on every bloody field equipment and anything above a 20mm grenade should be guided muinitions you have to make some compromises. And MEMS aren’t that bad yes they drift but for their usecases the drift is negligible.
MEMS operate on vibration which can be induced with solid state ultrasonics.
Also since most navigation systems today are hybrid drift isn’t much of an issue over long distances anymore and for short distance applications it doesn’t matter.
MEMS are much smaller, FOGS can also be made smaller especially considering you can use a single external laser injector to feed multiple sensors.
So while RLGS are still probably the most accurate production gyro we can make there are technical issues on making them smaller and since we have alternatives there isn’t likely much reason to invest in miniturizing them further if it’s even possible.
This is at least my educated guess.
The Thing was a bugged US seal built by the Soviet Union and given to the US ambassador to the Soviet Union as a gift.
What an interesting life you must have, I thought, if people gift you bugged aquatic mammals as pets...
Very likely the writers knew about the insectothopter, or came across it at some point.
People who would be interested in the misuse of this technology aren't, as a rule, interested in whatever laws you might think appropriate for regulating it.
Shouldn't that be 'entomopter', or just 'ornithopter'?
Looks like Merriam-Webster agrees:
> "an aircraft designed to derive its chief support and propulsion from flapping wings"
https://www.merriam-webster.com/dictionary/ornithopter
I wonder how long it will take the RC community to duplicate it?