Hacker News Reader

Comment by ChuckMcM | Hacker News Reader

This has been discussed before and is a favorite of some of the robotics communities. This particular drone seemed to be a bit overloaded (2.7kg?) and I'm guessing the person who launched it hadn't made the connection between flight time and weight. But still it seems a pretty straight forward way to get drugs from point A to point B.

That said, 2.7kg of crystal meth might be worth $270,000 [1]. So the cost of the drone is insignificant to the 'value' of the meth it was carrying. Sending 1kg chunks over on $10,000 drones would still net a decent profit. And $10,000 is a pretty nice drone.

[1] http://www.pbs.org/wgbh/pages/frontline/meth/faqs/

View on HN

VICE claims this was a "DJI Spreading Wings s900" drone [1]. There is a nice little demo video on youtube [2], but it is hard not to mentally replace "camera gear" with "6.5 pounds of meth", when watching it after reading the news story. Looks like it is about $3,800 USD rigged up [3]. The specs say it can only fly for 18 minutes [4], so they had to have been pretty close by! As battery tech, software auto pilots, and availability improves, this is likely going to become a much larger issue.

[1] https://news.vice.com/article/drone-carrying-three-kilos-of-...

[2] https://www.youtube.com/watch?v=8aC9Z7-Qxwk

[3] http://www.popphoto.com/gear/2014/08/new-gear-dji-spreading-...

[4] http://www.dji.com/product/spreading-wings-s900/spec

You can fly a pretty good distance in 18min @30-40mph (Napkin Math - somewhere in the 10mile one way range) it would be reasonable to speculate they were within 10km as they still had a way to go to get across the border but that's still a pretty large area.

You can quite easily build a drone capable of flying for many miles with a 2.7kg Payload for 2-3K. Indeed $10,000 would buy you a very, very nice setup. If they wanted to go the plane route a Mugin will do 15kg and still be under 10k [1]

[1] http://www.fpvmodel.com/super-huge-mugin-4450mm-uav-h-t-tail...

Yup, that would do it. 600km range for 15kg of 'cargo'. Assuming a full write-off of the air frame in the flight, possible recovery of the auto-pilot/electronics. It's right on the edge of what its radar profile is going to look like. (model/bird vs small aircraft).

Anyone know what that's based off? A "full sized" drone of maybe 3-4X this size, but with exactly this profile flew directly overhead at low altitude about 18 months back.

This perhaps? http://en.wikipedia.org/wiki/IAI_Heron-1

What benefit would these hexcopter drones offer over a similarly sized fixed wing drone? Isn't the safest smuggling flight for this a straight shot?

Copter might be easier to land in a discreet location. Does a fixed-wing require strip to land? You could always just crash it and eat the cost of the drone.

You wouldn't need much of a strip to land. A plane capable of carrying 6 pounds could land with a couple yards of short or medium length grass, and you could probably just dump it without doing damage beyond minor repairs.

Having flown many fixed wing and multi-rotors of sizes small to large fixed wing tends to require more skill to pilot and are harder to get autonomous landing right (a bad reading from the barometer on landing and you're toast) at least with the Pixhawk on a multi-rotor it'll stay hovering over the landing point until it is relatively certain that the GPS height and barometer readings match before landing. I've witnessed planes nosediving into the ground under the same conditions. Now indeed there are extra sensors that can help with this Lidar etc. but for a standard setup it's an issue.

Why land though? When you're talking this sort of money for the cargo, just drop the drugs and crash the thing, and write the airframe off as an expense.

I assume the fixed wing system would also be quieter.

Availability maybe? These copter drones are abundant now.

With a lot of the higher end drones, you can do autonomous flying via GPS. Just tell the drone a point on the map to go to and off it goes. You could easily have a partner in crime just waiting at the pre-determined GPS coordinates to pick up the drone and the cargo. The partner could then swap the battery pack and send the drone right back to you.

This can be accomplished right now, with pretty much any of the off-the-shelf, high-end consumer drones.

Setting up a fixed wing drone requires a lot more work. You would have to set up a parachute with the payload and some sort of release system, plus it would have to have enough battery to fly back. The receiving party on the other end wouldn't be able to send anything back with the drone, either, because it would never land. If you wanted it to land, then you have to deal with all sorts of other variables.

Drones that can take off and land vertically are far more flexible for point to point deliveries than fixed wing drones.

Fly back?

You'd probably just drop the cargo and crash the drone somewhere random nearby.

If you're smuggling 100k of drugs (or even 30k of drugs), you can just accept the 5k airframe as a transport cost and move on.

I'm pretty sure both the Pixhawk and Ardu Pilot Mega are capable of auto-landing a fixed-wing aircraft. Here's a demo: http://www.youtube.com/watch?v=5CF_vDq2Wq4

That's the street price. I'd ballpark guess that the wholesale price within drone-flying distance of the border is a tenth of that. Still profitable though.

In Australia it'd be about a fifth of the original figure, but our economics for the drug trade are very different I suppose.

I believe the closer to the border, the lower wholesale prices are.

Even a kilogram of marijuana would be somewhere in the neighborhood of $10,000. Something like pure cocaine is worth vastly more.

For this to work in the longer term, the drone would need to fly high enough to be more or less undetectable from the ground, and a long enough distance so that it's impossible to monitor all available landing sites. I don't know enough about drones or the US-Mexico border to say whether that's practical.

You mean low enough to the ground to be undetectable - and there is already hobby-level advanced technology to do this out on the intrawebs. I myself am building drones (planes, mostly) to do things like auto-navigate the neighborhood and so on, and am actively seeking ways I can get involved in a startup to use this technology for positive effects. I'm not at all surprised to see the drug cartels already making headway on carrying loads - there are a near infinite number of other good applications of this technology, which I sincerely hope governments allow to prosper, anyway, in spite of this ...

Your first price is about two times what top shelf, domestic black market prices are presently at.

http://en.wikipedia.org/wiki/Nap-of-the-earth

Flying low actually helps you against ground-based watchers, because they don't see you for very long as you zoom over them. You're on the ground and you detect a drone... and bam, it's gone.

so how long before

a) buying one puts you on a list

b) you are required a license to fly one, anywhere

c) you are required a license to buy, let alone fly one

I'd imagine that 3D printing is already good enough to print most of a drone; the only things you'd need to order would be wires, a control board (arduino?) and motors.

You usually need something strong for the frame; big drones are carbon fiber - and I don't think you can print anything like it right now.

Other things on the parts list are the motor controllers (motors are controlled by current, and you usually need at least 40Amps for each), battery and sensor board (GPS, compass, 3D acceleration and gyro).

There's no need for carbon fiber, and besides, you can buy carbon fiber beams and poles that are used for much more than just RC aircraft. Any multirotor frame you can build with carbon fiber you could build with fiberglass, aluminum, or even wood and plastic without significant performance reduction that would be relevant to hauling drugs short distances.

http://www.thingiverse.com/thing:267008

http://www.thingiverse.com/thing:171474

http://www.thingiverse.com/thing:175073

http://www.thingiverse.com/thing:241055

https://www.youmagine.com/designs/drone-race-w4-drw4

Well its been "possible" for a long time (over 10 years) using fixed wing craft. My favorite was this story: (http://www.washingtonpost.com/local/obituaries/model-airplan...) which talks about the guy who flew an RC model across the Atlantic in 2003.

That said, basic police work can still generally track the origin of a drone back to when it was built/sold. Every electronic component has a date code, and there are records of sales all along the way. That can tie it to a location/purchase, in store surviellance can often provide images of buyers, (mail/internet order is even easier) and get back to who had the drone to begin with.

At the end of the day, the higher the tech 'toys' used in a crime the easier it is for law enforcement to figure out who did it.

Every electronic component has a date code, and there are records of sales all along the way.

I have a feeling that if you buy your electronic components this way:

http://www.jameco.com/webapp/wcs/stores/servlet/StoreCatalog...

it's a bit harder to trace an individual component back to the purchaser. Especially something like this:

http://www.jameco.com/webapp/wcs/stores/servlet/Product_1000...

No doubt. And if you were really trying to be stealthy you would open up an "e-waste" store and use it to collect things which you could scavenge parts out of. (would make for some good fiction as well). Thinks like GPS units are hard to get in junk bags, but easy to pick up at a recycling store out of an old phone (assuming it isn't part of the SOC). But the point remains that all of those things have 'tendrils' and the mosaic effect is very real.

Most of what you need to build an autonomous air vehicle is already in your phone: https://code.google.com/p/andro-copter What are you going to do, restrict brushless DC motors, LiPo batteries, carbon fiber tubing / ABS plastic, and small propellers (they still use wooden ones on some planes, which would also work on quads) ?

Please keep in mind, the USA is not the only supplier of these parts/components.

GL tracing a drone built from part kits bought of deal extreme or Ebay.

Most of those drones are built from general purpose electronics which even if they were original and no knock off which is very common these days they would've been bought from some reseller in bulk and no one actually tracks individual components.

While i agree that it's easy enough for a detective to go and ask around who has a drone but once you have enough of them flying around that won't be as simple either.

I think the big change has been in the autonomous navigation combined with the pre-assembled stuff that is really hard to crash unless you're trying. For a few grand now, you can get a more or less plug-and-play (or charge-and-fly) system capable of transporting a few pounds of cargo a dozen miles with a minimally trained operator and minimal assembly.

You could also use a big model rocket, but you can't buy a huge one off the shelf that works so reliably with little skill. RC subs would also be possible, but again, you'd need to do a lot of your own R&D, whereas companies like DJI now do it for you. Big RC helis have been capable of the same-ish payloads for a decade or more now, but they're hard to fly, hard to build, and require quite a bit of training to fly without crashing. You'd also need to be watching it every second in flight, unlike the self-stabilizing GPS waypoint systems which have become affordable (<$10k and available to non-military) in the past few years.

You can always build a multicopter with parts that are not specifically made for drones, so they cannot enforce this. And when you deal drugs, smuggling a drone is the easy part.

As someone who recently picked up the multi-copter hobby, I would say I want people to have to register/license these things. Previously I might have scoffed at the "flying lawnmower" meme, but it is absolutely the truth. Even the small 250mm mini sized quads are massively dangerous and the larger octocopters are downright scary. They move fast and 10" carbon fiber blades spinning at 8000rpm connected to a 5lbs drone could all but decapitate a human. One GPS glitch and the thing might go full speed toward your face, or a kid on a bike, or the neighbors dog / beamer / house. How much do you trust GPS and hobbyist open source flight control code? I trust it enough to fly around an empty park, but I'd never want it near my cat.

With the prices for entry level models coming low enough to compete with an ipad or other consumer electronics, I can't help but think how scared I would be of teenagers flying these things around my neighborhood, let alone the hazards with wide scale deployment of automated drones for deliveries.

My favorite battlebots story was from a teen who built a 'spinner' and early in their testing they lost a nut when the blade was in full spin. It left a hole in the siding of his garage as if someone had shot a bullet out of the house, 14" off the floor. Fortunately nobody was hurt. He did all future tests in a drained swimming pool.

Jamie Hyneman and Adam Savage had their robot Blendo removed from competition twice because it was too dangerous.

https://en.wikipedia.org/wiki/Blendo

This is just so awesome.

> early in their testing they lost a nut

I'm glad I read this till the end. For a moment there I was thinking the kid had actually lost, you know, one of his nuts.

I've been in the hobby for almost 2 years, and my impression is very nearly the opposite of yours. I think the danger of hobby-grade quads is almost universally overstated. They're dangerous, but not significantly more dangerous than kids playing baseball, and certainly nowhere near as dangerous as an automobile. I think your fears are unfounded.

Less dangerous than a car? Perhaps. Although I'd point out that we require licensure and insurance for cars, and the threat of life in prison if you injure anyone through neglect while operating a vehicle

Baseball? No. These are injuries from 450mm sized quads.[1][2][3]

warning some blood / injuries

[1] https://www.youtube.com/watch?v=ji3Hii_LZOc#t=20

[2] http://static.rcgroups.net/forums/attachments/7/5/2/8/2/a300...

[3] https://www.youtube.com/watch?v=ocqB6_y71xE#t=6

Yes, spinning rotors can cut your skin if they make contact. That has to be weighed against the likelihood of spinning rotors and skin making contact. That likelihood is generally fairly low, but is much higher when a pilot deliberately flies straight at a person's head or throttles up the aircraft while holding it.

Baseball tends not to cause lacerations, so photographs won't look as gory (except perhaps for some eye injuries), but they definitely can cause serious injuries, and can very easily cause minor property damage.

A 700-sized heli can kill you, but I'm not sure that even a large drone would do as much damage: http://www.dailymail.co.uk/news/article-2413231/Roman-Piroze...

I say this because the heli in the story, as most high-end consumer RC helis these days, had 2 carbon fiber blades, and on the 700 size, tip to tip on the blades is about a meter. Consider the weight and tip speed of each blade vs. A large quad or hex or octocopter. The blades will be much smaller, so much much less kinetic energy at the tips. You also don't do this with quads: http://youtu.be/EYWLqui5VsM

No, but you can do this with quads:

http://vimeo.com/68798774

https://www.youtube.com/watch?v=f5eUqLopMYM&index=88&list=PL...

I have multiple(big,expensive) drones. I program them too.

I am not too worried about them, they are far safer than a real helicopter.

"One GPS glitch" ?? When something fails on a drone you could detect it very easily. If GPS fails you have barometric sensors, you have accelerometers, you have compass, you have gyroscopes, you have infrared sensors for horizon.

When something fails you actually reverse and disconnect the motors automatically. Now it is just 1 or 2 kilos falling down. Mine have small parachutes.

Like humans you could add redundancy to those machines, and the probability of all of them failing at the same time is almost zero.

Normal toy quadcopters weight 100 grams and you can stop the blades with your hands (or you neck) without harm.

you need a licence and permission to fly one in Spain already

How would they enforce this without tracking everything larger than a small bird via radar or satellite?

i dont know how they enforce it but its a law. Spanish law works different from most countrys. They have laws to say what you can do rather than what you cant. The law is you need a licence.

With the number of idiots that have already almost collided with commercial aircraft, I think the drug connection is the least of our worries.

http://www.washingtonpost.com/world/national-security/near-c...

This reminds of South Park episode on the drones: http://southpark.cc.com/full-episodes/s18e05-the-magic-bush

Reply on news.ycombinator.com