> Drones can fly either high enough to not be easily seen
By eye, yes. By digital camera? That would have to be pretty high. At least on clear days. But there are days where visibility is low in general, and that goes for cars too.
> Or they can be launched near a target and fly low.
High risk targets should have additional safeguards. It's not hard for cars to get to some locations just because they aren't allowed, but because there is additional security and physical obstructions. One solution never covers all cases, and we shouldn't expect it to.
> But you need another licensed vehicle at that point to transport it.
Or you hire someone to tow it. Yes, it's harder. But also it's not like every time you get behind the wheel of a car someone checks that you're licensed. How many times have you had your license checked while driving? How many hours have you driven?
> You think that a 2 pound drone is only "slightly easier" to transport surreptitiously than a 2 ton car?
I think in the case you are planning to use it illegally, it's not really that much of a difference.
Seriously, you can probably walk into any rundown part of town, buy some older beat up car for cash for under $1000 dollars, and immediately use it for your purpose while the title change is in the mail (for states that allow that), if the person selling it really even cares.
Drones are new, so there's less infrastructure to track and manage them, but it's not really all that different IMO, it's just a matter of time.
> By eye, yes. By digital camera? That would have to be pretty high. At least on clear days.
Have you ever seen a hobby drone flown? I challenge you to see a typical model flying at 250ft feet with anything not using a telephoto lens, even on a clear day, unless you already knew where it was (aka saw it from a lower altitude and followed it up).
> I challenge you to see a typical model flying at 250ft feet with anything not using a telephoto lens, even on a clear day, unless you already knew where it was
Do you think you could identity a drone moving at 50 feet using a video camera that recorded as 640x480? Because approximately the same number of pixels would by used on a 4k capable video camera at 250 feet, and that's before any optical zoom at all. As long as you can identify something moving, it can be zoomed in on and classified (either by the same camera or a secondary one dedicated to the task), or just give the camera more pixels and use digital zoom. A few of those in a cluster would cover an area quite well I think.
For another way to think of this, an object that is 8" (assuming a 12" drone at an angle) should appear to be a quarter the size of the moon to the naked eye at 300 feet.[1] I don't think that's something a good quality camera should have trouble picking up.
1: http://www.1728.org/angsize.htm, input angle in minutes 7.5, size .666, using their example which says the moon appears to be about 40 minutes in angle size.
> For another way to think of this, an object that is 8" (assuming a 12" drone at an angle) should appear to be a quarter the size of the moon to the naked eye at 300 feet.
If a drone was a solid spherical mass like the moon, your guesstimate might have some merit, but it’s not. In reality, your 8” is already way overshooting it. Furthermore, the amount of light reflecting by the moon is significantly higher providing much higher contrast for edge detection. The matte white plastic of many common drones however easily blends into the sky.
Am I saying no camera would do it? No, that’s obviously wrong. But the post I replied to used a very vague wording which implies a majority of cameras would do it, which is definitely not the case. Happy to provide real photos to inspect this weekend (assuming it’s not still raining) at 100, 200, 300, and 350ft at whatever megapixel you’d like (up to 22mp) and focal length (from 8-1600mm, 35mm equivalent) so you can see what I mean.
> In reality, your 8” is already way overshooting it.
Given the body area of most these drones, I'm not so sure. I'd best the visible cross-section is pretty close to 64 sq. inches from most angles, if you include the small amount of area from the arms, but ignoring rotors.
> But the post I replied to used a very vague wording which implies a majority of cameras would do it
I don't read what I said that way. I said a digital camera, as in there exists a camera or type of camera that would work. I mean, there's some pretty crappy digital cameras out there, if you want to go really cheap.
> Happy to provide real photos to inspect this weekend (assuming it’s not still raining) at 100, 200, 300, and 350ft at whatever megapixel you’d like (up to 22mp) and focal length (from 8-1600mm, 35mm equivalent) so you can see what I mean.
Sure, I would be interested in see how it looks in practice. I purposefully limited myself to a little over 8 megapixels (4k) as I assumed video, and that's fairly easy to get with consumer grade hardware, which sets a nice precedent. If cheap consumer grade hardware can do it, a government solution with possibly special purpose hardware would have no problem.
Knowing how a decent still camera views it at full res with little or no zoom would be a good way to look at is, as I think that would approximate a high-end video system, or a mid-range one in a couple years.
I get emails on replies, so I shouldn't miss it even if it's a few days later.
>Should appear to be a quarter the size of the moon to the naked eye at 300
The moon is illuminated by the sun, so not really a fair comparison. Take a drone and paint it the color of the sky and it's going to be a lot more difficult than “a quarter the size of the moon” would imply. A drone can also fly much higher than 300 feet, and it's not a solid object.
But I've written a fair bit of object detection code though, and I'm certain a decent camera could do it fairly accurately.
False positives are the problem though. Birds, airplanes (farther away), insects (closer to the camera), clouds. All of those things are much more common than drones and so even a very small false positive percentage is going to cause this system to be completely unreliable.
That’s all without even mentioning licensed drones. How will the system handle not sounding the alarm when it detects a licensed drone? All drones will have transponders I assume? So now the camera needs a radio, and access to a national database of transponder codes.
Assuming drones become more popular how does the camera know which transponder code belongs to which drone? Signal strength? This will be different depending on environment. Signal direction? You need multiple receivers for this.
How do you stop a drone from emulating another drone’s transponder code? A challenge response system with a public/private key pair?
And the most important question yet. What do you do when you spot an unauthorized drone? If a cop spots a car he can chase it.
Does each camera have an interceptor drone?
How does the interceptor drone bring down the unauthorized drone without causing any collateral damage? What happens if there are multiple unauthorized drones in an area? How do you trace the unauthorized drone back to the owner? Unlike a car, the operator isn't in the vehicle.
Do you just allow the cameras to hand off monitoring to other nearby cameras until it lands and send cops to deal with it? How do you build a camera network with enough cameras to provide continuous monitoring, taking into account trees and buildings.
All of that just to say yeah it’s possible, but it’s a lot harder and a lot more expensive than it seems. Much too hard and much too expensive to feasible anytime soon. Unless drones prove to be a lot more popular and effective than truck, car, and suitcase bombs, than I think they will, no one is going to pay for all of this.
That’s not to say that we won’t have more licensing and registration for drones, but I don’t think it’s going to prove very enforceable or very effective at anything more than reducing the number of nuisance complaints. Maybe it’s worth it just to do that, but I’d need to see real data first.
That's also not to say that some kind of automated drone defense won't almost certainly become more popular in very specific high security areas like the White House.
> The moon is illuminated by the sun, so not really a fair comparison. Take a drone and paint it the color of the sky and it's going to be a lot more difficult than “a quarter the size of the moon” would imply. A drone can also fly much higher than 300 feet, and it's not a solid object.
A good point. I think flying above 400 ft is already illegal though, so dealing with that might be somewhat simplified.
> All drones will have transponders I assume?
I think that might be a fair assumption. We require license plates for vehicles on the road, requiring transponders for use in populated areas sounds like a good way to track and license (if your transponder is issued, there you go).
> So now the camera needs a radio, and access to a national database of transponder codes.
Well, yeah. If it's a system responsible for tracking and ensuring licensing compliance and flight permission, it would, just as bridge toll systems now often use license plates to track fees. It's not all that hard in this day and age.
> Assuming drones become more popular how does the camera know which transponder code belongs to which drone? Signal strength? This will be different depending on environment. Signal direction? You need multiple receivers for this.
You already likely need multiple systems to get good visual and radio coverage. But even then, it doesn't need to be a perfectly self-contained system, just a piece of the puzzle. If you see six objects and receive 5 signals, eliminate the ones you can, and send someone to investigate with directional equipment.
We don't expect red-light cameras to pull us over, but we do expect they do something. It's not crazy to expect they also might have specific circumstances that flag to additional review for possible dispatch of an officer.
> How do you stop a drone from emulating another drone’s transponder code? A challenge response system with a public/private key pair?
How do you keep one car from using another's license plate? You make the sale of fake license plates illegal, you make sure the only valid plates are issued by the government, and you make it illegal to tamper with them.
> And the most important question yet. What do you do when you spot an unauthorized drone? If a cop spots a car he can chase it.
Already covered a little above, but there's already solutions appearing to help with this. There's anti-drone ammunition[1] and trained animals[2] on the simple end, up to military grade anti-drone solutions.
> How does the interceptor drone bring down the unauthorized drone without causing any collateral damage? What happens if there are multiple unauthorized drones in an area? How do you trace the unauthorized drone back to the owner? Unlike a car, the operator isn't in the vehicle.
What do we do with vehicles or trailers without clearly defined ownership we find on the street? It's towed to an impound lot. If someone wants to claim property, they they come for it.
If downing it for legitimate reasons ensues, the owner of the drone would be responsible for damage, and if the owner can't be determined, I'm sure there's a legal precedent to fall back on. It's not that wild a situation.
> All of that just to say yeah it’s possible, but it’s a lot harder and a lot more expensive than it seems. Much too hard and much too expensive to feasible anytime soon. Unless drones prove to be a lot more popular and effective than truck, car, and suitcase bombs, than I think they will, no one is going to pay for all of this.
I don't think it's nearly that expensive, as outlined above. If there isn't something like this, I expect drone terrorism would definitely be the go-to solution. This wouldn't stop that, but it does put a hurdle in the way (in the same way that doors and locks and security systems put a hurdle in the way of a home invader, but in the end only slow and dissuade, and do not actually prevent in most cases).
> I think flying above 400 ft is already illegal though
So is using a drone for a terrorist act, so by that logic, wouldn’t that negate the need for this system you propose?
> I don't think it's nearly that expensive, as outlined above.
To cover what area? If you mean something like the White House, sure, the cost would very likely be justified.
If you are talking covering the entire United States, or even just everything important, you are insane (not meant to be rude, but a trillion dollars wouldn’t be unrealistic estimate there).
I think a key point you are missing is what real threat a consumer grade drone is in reality.
$500 Drone - might be able to lift a hand grenade. Keyword there might.
$1500 Drone - the hand grenade is now pretty solidly possible depending on the type of grenade (M67, doable, but anything bigger less likely). In the best case you might get up to a pound of cargo weight, but it’s significantly going to impact maneuverability and flight time.
$10k+ drone - once we get into this category, sure, we can get up to 5lbs or maybe slightly higher in cargo weight, but then we are in the category of 4ft or larger hexacopters or octocopters. They are very noticeable and very loud.
I think that’s the real reason we’ve not seen much (and likely won’t see much unless drones dramatically change in lifting capacity) terrorism from drones. To put it bluntly, there are just too many ways to deliver tiny explosives that are more practical and less likely to fail than are with drones. For the realistic payload of something like an M67 grenade, the “kill radius” is only 5 meters. If I saw a drone with something attached to it (or even if not) within 5m of me, I’m already going to be getting out of the area. Sure, you could argue someone may intend to drop it from altitude, but even then do you really think the accuracy of dropping something like that is actually going to be very precise?
The terrorism angle is in reality pure FUD and is used as an excuse to appeal to the uninformed masses.
> So is using a drone for a terrorist act, so by that logic, wouldn’t that negate the need for this system you propose?
No, I'm just noting that however we deal with this for aircraft might help here, since the same rules apply.
> To cover what area? If you mean something like the White House, sure, the cost would very likely be justified.
High risk locations. Dense metro downtowns, sports stadiums, government buildings, large public utilities (nuclear power plants), etc.
> If you are talking covering the entire United States, or even just everything important, you are insane (not meant to be rude, but a trillion dollars wouldn’t be unrealistic estimate there).
Assuming 10,000 locations across the U.S., and assuming 50 $50,000 average installation cost (which might be low, but we can scale it easily for ballpark costs after), that's $500 million. If we double locations and quadruple average install cost (which is supposed to account for locations with multiple towers, etc), that's still just $8 billion. That's totally doable when it comes to stuff tagged as national defense.
> I think a key point you are missing is what real threat a consumer grade drone is in reality.
Your prices and payloads are way off. A few grand ($3k-$6k) gets you 20 lbs of payload easily (and more if you need it), from multiple manufacturers.[1] $10k seems to get you like 40 lbs or more.
> For the realistic payload of something like an M67 grenade, the “kill radius” is only 5 meters.
Grenades are not optimized for damage. They are a compromise that takes into account explosive power, compact size, and delivery. The fact they are meant to be hand delivered to within a few tens of feet means a large blast radius is actually detrimental to their design. 10+ lbs of payload for something that would want the maximum blast radius and power, or the maximum directed penetration for hard targets, probably allows for quite a different set of compromises.
> The terrorism angle is in reality pure FUD and is used as an excuse to appeal to the uninformed masses.
There's a lot of FUD, but I don't think it's complete FUD. It's a real attack vector, and if there's no mitigations, I think it's possible it will have the best cost to benefit ratio (since the chances of finding the perpetrator would be virtually nil, and the cost is relatively low), and that's all it takes for it to be used.
>High risk locations. Dense metro downtowns, sports stadiums, government buildings, large public utilities (nuclear power plants), etc.
You take out dense metro downtowns and I think that's perfectly sensible.
It doesn't require any kind of licensing, or national database, or wide scale distributed systems of any kind. Drones are off limits in the those areas just like aircraft are. If you have an automated defense system and you need a drone for some exceptional reason, you clear it with whoever manages the local system and they install a friend or foe transponder on the drone.
Wide scale drone licensing enforcement isn't required or effective if your goal is to protect high threat targets from drone attacks. It solves completely different problems.
>No, I'm just noting that however we deal with this for aircraft might help here, since the same rules apply.
You need a much higher density of radar installations, with an exponentially higher cost since drones are much smaller than aircraft.
>$8 billion
That's off by a few orders of magnitude if you're talking about a wide-scale system including towers, radar, personnel, drone take down capability etc...
>10+ lbs of payload for something that would want the maximum blast radius and power, or the maximum directed penetration for hard targets
You’re still talking pipe bomb levels. It’s still much more effective to leave a bag in a trashcan in a crowded space.
> A few grand ($3k-$6k) gets you 20 lbs of payload easily (and more if you need it), from multiple manufacturers.[1] $10k seems to get you like 40 lbs or more.
You're looking at a list of maximum payloads from the marketing department. Not maximum realistic operational loads. And the drones with 40lb payloads are huge.
>It's a real attack vector, and if there's no mitigations,
There are no feasible large scale mitigations that are effective enough to change the calculus.
Again if you’re talking local mostly automated drone defense, then sure that’s feasible and probably necessary at some point.
I'm not feeling this is a particularly constructive discussion anymore. You seem to be picking specific statements to counter, while ignoring the context of the statements and explanations that address your counterargument I've made in a prior comment or within the same comment.
For example, drone payload, where you focus on smaller sizes, and say they can't do much damage, and then either ignore larger size payloads as marketing lies (with no evidence to back it up, or to at least show actual payloads in practice), or that those drones are huge... which would make them easier to track, which was one of my earlier points that makes the system easier to develop. Feasible drones for attacks will likely be larger, and this easier to detect.
Time an again, I've tried to come up with real data or approximate numbers, with all calculations and reasons for such outlined so you or anyone else could question where you thought my assumptions were wrong, but I haven't seen that. Instead, I get blanket rejections without evidence, or even much explanation of reasoning.
I think you're conflating 2 separate people's statements. I've never had any focus on smaller sized payloads.
You could fairly easily find a drone with a 10 lbs payload. 40 pounds though isn't a realistic operational payload for those drones, which is why it says "max payload".
>Feasible drones for attacks will likely be larger, and this easier to detect
That's true, my point is that setting up a nationwide drone licensing and tracking system to track hobby drones isn't necessary when what you're really worried about is protecting high value targets from drones that are approaching the size of small helicopters.
> Instead, I get blanket rejections without evidence, or even much explanation of reasoning.
You haven't provided any evidence either, just a random assumption that such a system will cost $50k-$200k per site.
I clearly explained why that is too low. Increasing the density of radar coverage to handle drones above 400', trained personnel, towers, and drone take down systems are going to cost a lot more than you think. You can believe me or not, but I've provided at least much explanation and evidence as you have.
> I think you're conflating 2 separate people's statements. I've never had any focus on smaller sized payloads.
I know you weren't the prior commentor. This is a long discussion. I expect you're reading the prior comments in the same direct thread. When I've specifically addressed an item, in the thread, or even in the same comment, to have that ignored is... frustrating.
> You could fairly easily find a drone with a 10 lbs payload. 40 pounds though isn't a realistic operational payload for those drones, which is why it says "max payload".
I didn't pull that number out of my ass. It was actually a conservative number, based on what I saw some models as capable of (but could not find good pricing info for). For example, the VulcaUAV Airlist model advertises a payload capacity of 25kg (~55lbs).[1]
> That's true, my point is that setting up a nationwide drone licensing and tracking system to track hobby drones isn't necessary when what you're really worried about is protecting high value targets from drones that are approaching the size of small helicopters.
Not just hobby drones, all drones. I expect there to to be a lot of commercial drone usage in the future. If they are all licensed, then the costs are limited to applying towards the small hobby usage, but overall U.S. drone usage, which I expect to be substantial compared to hobby use.
I've explained this multiple times.
> You haven't provided any evidence either, just a random assumption that such a system will cost $50k-$200k per site.
I've explained what type of hardware I expect the system to require, including multiple cameras per station. I've given a guesstimate as to the number of installations required. I've guessed at a per-station cost, and thus a total system cost.
Your counter of "it's off by a few orders of magnitude" is in my opinion not sufficiently explained by personnel and countermeasures, since a few order of magnitude is at a minimum 100x off (and then if I assume you used a meaning of "few" that includes 2 as the minimum). I do not believe training existing personnel that are already stationed around every location (police!) you would want these and providing additional devices and training constitutes hundreds of billions of dollars.
> I've provided at least much explanation and evidence as you have.
No, you haven't provided any idea of what you think it would actually cost to have personnel on site and have them trained and outfitted. If you had, I might have been able to point out that we already have officials that serve a very similar purpose that could be used, the police, and that I don't think training and outfitting would be nearly as costly. I think $100 billion would be an fairly extreme upper bound of this system, but that's not a few orders of magnitude, it's one, and surmountable over time with a tiered rollout IMO, if it's event that high. For example, the FAA alone out of all the DOT agencies has a $16 billion annual budget.[2] In 2018 the U.S. spent over $90 billion on transportation in the federal budget.[3]
In the end, I doubt we're going to convince each other of anything at this point, and I've lost the desire to continue. We should just agree to disagree. :/
Feel free to address any of the points I've made here though. I'll read the response and consider the points, I just don't think I'll respond.
>I didn't pull that number out of my ass. It was actually a conservative number, based on what I saw some models as capable of
I read your earlier source. And I'm telling you that a list of drone maximum payloads pulled from the marketing websites of mostly small businesses isn't representative of the actual normal operating payloads of drones at those price points.
Most drones can technically take off with a much higher weight than is actually practical to fly with.
I'm not saying you can't find a gas powered drone with a 100 pound payload. What I am saying is that these drones aren’t remotely in the same category as consumer drones.
>Not just hobby drones, all drones. I expect there to to be a lot of commercial drone usage in the future. If they are all licensed, then the costs are limited to applying towards the small hobby usage, but overall U.S. drone usage, which I expect to be substantial compared to hobby use.
Then what is the point in regulating tiny drones with little capability to cause damage. Ultralight aircraft aren’t regulated like commercial jets.
>Your counter of "it's off by a few orders of magnitude" is in my opinion not sufficiently explained by personnel and countermeasures, since a few order of magnitude is at a minimum 100x off
This depends on what you’re trying to do. If your goal is just to make sure that commercial drones are licensed, then sure you could build a relatively cheap system. If your goal is just:
Hey we keep seeing an unlicensed drone delivering pizzas. Alert the police to go out with direction finding equipment when they get the time, and see if they can find the operator, so that we can fine him.
Then sure a few billion is doable. Red light cameras don’t deter terrorist from using truck bombs though. If we removed the requirement to have license plates tomorrow, we wouldn’t see a huge upswing in truck bombings. Stealing a license plate isn’t the bottleneck in the terrorists plan.
So if your goal is to meaningfully deter determined attackers, you’ll need a system that’s much more robust. You’ll need enough extra trained personnel to rapidly respond to a relatively large percentage of every unauthorized drone alarm. You’ll need a very dense network of anti drone countermeasures (birds, guns whatever) in order to have a chance of responding in time. You’ll need vastly expanded radar coverage to detect drones flying too high to be detected by camera. And a camera network that is dense enough so that just flying behind building and trees isn’t an easy option. All of this will need to coordinated.
>In 2018 the U.S. spent over $90 billion on transportation in the federal budget.
That's in one year, and the DOT doesn’t operate anything close to the scale of the system we’re talking about.
> You’ll need vastly expanded radar coverage to detect drones flying too high to be detected by camera.
I’d actually argue it’s not possible with current radar tech. The radar signature of most consumer drones is so small that if you had radar tuned to that size, you’d be getting false positives all the time. Even in the bigger drones, there’s still a very small radar signature too. I’ve not tested it, but even my 6’ wide dual octocopter likely wouldn’t even be identified on radar.
> Your prices and payloads are way off. A few grand ($3k-$6k) gets you 20 lbs of payload easily (and more if you need it), from multiple manufacturers.[1] $10k seems to get you like 40 lbs or more.
Actually, my numbers aren’t off (and the companies aren’t lying in their marketing directly), it’s simply a misunderstanding over the units of scale.
For purposes of discussion, I’m going to focus on the payload numbers used in the link you provided, not manufacturer spec sheets as they vary a lot in terminology. I also own three of the drones listed in that link too, so what I’m saying isn’t hypothetical assumptions, but real world experience.
Let’s also use the DJI S1000 (but I’m happy to break it down for others too) for this discussion. The article claims a payload capacity of 7kg. This is actually a bit off flat out as the max takeoff weight is 11kg and the drone in question weighs 4.4kg. So our number is actually 6.6kg for payload. The biggest disconnect though is that the 4.4kg is just the airframe / motors / props / mcu, it doesn’t include the batteries (biggest impact) or some other needed electronics. Unlike with your average consumer drone, batteries are something you have wide control over in size, weight, and capacity. That said, unless you want to only fly a very short distance (think short walking range), you are still stuck with a heavy pack. On the low end, expect about 2kg, on the high end 3kg. That puts us under 4kg of payload now. If we want to see where it’s going (allowing us to be further away from this hypothetical bomb) or have more telemetry, we are going to be getting closer to the 3kg payload mark. Now let’s talk about price. These big drones are money pits. Sure, you see those $3-6k price tags advertised, but the part you don’t realize until you get into the field is that’s not a “flight ready” price. You are going to need batteries, radios, battery chargers, etc etc to actually be able to fly. For these big drones, you’ll spend >$500 just for one flight’s worth of batteries if done right. Shockingly, for a good battery charger, you’ll spend nearly a grand. For an appropriate radio, $500. You want remote video too? That’s another $1-2k, not including a gyro or camera. Heck, I spent pretty close to $1000 just on the carrying case for my S1000. Yes, a terrorist would likely cut a lot of corners, but it’s still going to be a lot more than the base sticker price you see advertised and my number of $10k is a fair median point for what it’d really cost on these bigger drones. Could you do it cheaper, yes, but my numbers are still in the right ballpark.
—
I agree with your later point to another person, I don’t think this venue is encouraging a healthy discussion. You come off as someone intelligent, just not personally experienced in this field. As such, I’d be more than happy to continue the discussion offline. Shoot me an email if interested, info in my profile.
> it doesn’t include the batteries (biggest impact)
I suspected this (it wasn't spelled out, but became likely as I was researching). That said, I think it's somewhat mitigated by the fact long flights wouldn't really be necessary. Enough battery to fly the route and perform the end action, with some sufficient overhead (an additional 50-100% capacity?) would possibly be sufficient. A few minutes of flight time likely gets you quite far.
That said, it also works towards my point of specific areas that are tracked may be sufficient. If it's hard to fly a drone from miles away to avoid launch site detection and still have a delivery payload large enough to be useful, that's beneficial in itself.
> my number of $10k is a fair median point for what it’d really cost on these bigger drones.
Fair enough. I agree that actual deployment is probably either an additional $5k-$10 or some multiple of the base price, depending on how some items scale.
It leaves the question open as to whether the larger drones could still be useful though. Being able to fly a bomb up to a corner office in some building and detonate it, or quickly drop down to a motorcade someone has entered, might be a good return on investment even at $30 or more, depending on the target.
> just not personally experienced in this field
That's correct, and it's why I value your input on this, since you seem to have a lot of experience with the actual drone portion.
I don't mind being proven wrong, or shown evidence I'm wrong or off in calculations. I do mind just being told in wrong without new number or without specific objections provided as to my methodology, since I went through the effort to provide them in the first place. As such, I appreciate you providing some real numbers to specific models for this discussion.
> I’d be more than happy to continue the discussion offline.
I appreciate the offer (I really do!), but I'm worn out by this discussion at this point.
Additionally, I doubt we'll make much headway on the actual tracking system part of the discussion, there's too many variables, and too much variation in our different ideas of what those variables entail. As those variables stack and the differences multiply, the expected end values diverge too wildly.
Since I think neither of us have enough hard data to reduce enough of those variables to fairly static values without quite a bit of research on each (such as how many priority targets are there, how the military treats this topic, current state of the art video tracking techniques, the effect of weather on video tracking, etc, and truthfully that's a lot to invest in an internet discussion), it seems unlikely to be all that fruitful. :/
You want to spend hundreds of billions of dollars on a national system to counteract the threat of determined attackers that is ineffective against determined attackers?
Building an effective IED that can be transported by a drone is already much harder than bypassing any realistic national anti drone defense. Who is this system mean to stop?
>A good point. I think flying above 400 ft is already illegal though, so dealing with that might be somewhat simplified.
We are talking about determined attackers, so it being illegal isn't really relevant. Unless you want to cover the country in radar that can track anything the size of a drone that’s above 400’, so the camera system on has to handle below 400’. That’s possible, but now we’re adding hundreds of billions more to the system.
>just as bridge toll systems now often use license plates to track fees. It's not all that hard in this day and age.
Bridge toll systems are nothing like the proposed system. There is a huge difference between building a camera that takes pictures of license plate that sends a ticket in the mail weeks later, and a distributed system with hundreds of thousands of access points that each need to access a national database to make decisions in real time.
>We don't expect red-light cameras to pull us over, but we do expect they do something. It's not crazy to expect they also might have specific circumstances that flag to additional review for possible dispatch of an officer.
Red light cameras don’t do anything like this at all. The amount of false positives is far too high. There is a reason you get a ticket in the mail weeks later.
> If you see six objects and receive 5 signals, eliminate the ones you can, and send someone to investigate with directional equipment.
Drone flight times are usually under 20 min. Even in the future with battery technology, it’s unlikely someone is still going to be operating by the time you get someone out there with directional equipment. And now we have a to train a nationwide drone force capable of operating directional equipment.
>You make the sale of fake license plates illegal, you make sure the only valid plates are issued by the government, and you make it illegal to tamper with them.
None of these will dissuade a determined attacker. Just like license plates don’t stop terrorists. They aren’t designed to do that.
>Already covered a little above, but there's already solutions appearing to help with this. There's anti-drone ammunition[1] and trained animals[2] on the simple end, up to military grade anti-drone solutions.
So anytime a transponder malfunctions you want to shoot at a drone over a populated area? You want to want to add gun turrets to the cameras covering the country? Or do you want a nationwide system of trained attack birds stationed every mile or so?
You are massively underestimating the difficulty to accomplish what you are proposing.
The amount of explosives a drone can carry mean that for the foreseeable future they aren’t going to be particularly effective weapons for terrorists. Certainly not effective enough for us to spend hundreds of billions of dollars on a system that can be easily bypassed by flying below the tree line.
Yes, I agree automobile regulations won't stop a determined terrorist from using a car to commit a crime.
I think you've done a great job illustrating why similar laws requiring licensing and registration will be completely ineffective at stopping terrorist from using drones, which is the context of this whole discussion.
By eye, yes. By digital camera? That would have to be pretty high. At least on clear days. But there are days where visibility is low in general, and that goes for cars too.
> Or they can be launched near a target and fly low.
High risk targets should have additional safeguards. It's not hard for cars to get to some locations just because they aren't allowed, but because there is additional security and physical obstructions. One solution never covers all cases, and we shouldn't expect it to.
> But you need another licensed vehicle at that point to transport it.
Or you hire someone to tow it. Yes, it's harder. But also it's not like every time you get behind the wheel of a car someone checks that you're licensed. How many times have you had your license checked while driving? How many hours have you driven?
> You think that a 2 pound drone is only "slightly easier" to transport surreptitiously than a 2 ton car?
I think in the case you are planning to use it illegally, it's not really that much of a difference.
Seriously, you can probably walk into any rundown part of town, buy some older beat up car for cash for under $1000 dollars, and immediately use it for your purpose while the title change is in the mail (for states that allow that), if the person selling it really even cares.
Drones are new, so there's less infrastructure to track and manage them, but it's not really all that different IMO, it's just a matter of time.