Comment by swalsh | Hacker News Reader

Comment by swalsh | Hacker News Reader

People say that there's usually about a 10 year lag before tech from the military enters civilian use. Since drones showed up roughly 2003, it stands to reason that 2013 is when we should first starting seeing the tech become civilian viable. In fact that seems to be true, there's a whole lot of DIY material out there, and that's from people who don't have real world experience. I wouldn't be surprised if as military contracts wind down the large amount of people who just developed a new set of expertise may try to break out into the private sector. The only thing stopping the tech seems to be the FAA. Obviously its not something to take lightly, there are real safety concerns here... but its not completely science fiction.

Why do we accept that modern warfare can be fought almost entirely with drones (libya?) but when it comes to civilian applications it's a fairy tail publicity stunt?

I think bezos wants to turn a profit, and lowering delivery costs, and giving faster service might only make his company that much more valuable. I don't doubt for a second that this is a genuine effort.

These drones owe nothing to military tech.

They're born out of the hobbyist RC scene and were enabled by Chinese investment in cheap RC parts development.

Yeah... that's what my post said, but there's obviously some common ground between them, and my speculation is that reduced military funding might encourage people who have that expertise to branch out into the private sector.

2003?

First of all, I have no idea why people insist on calling UAVs (unmanned aerial vehicles) drones. It's a stupid term. Secondly, UAVs are much much much older than 2003.

http://en.wikipedia.org/wiki/Unmanned_aerial_vehicle

The United States was flying UAVs in the friggin 60's

http://en.wikipedia.org/wiki/Ryan_Model_147

http://gizmodo.com/the-ryan-firebee-grandfather-to-the-moder...

The term 'drone' long precedes the Pentagonese moniker 'UAV'.

I think it originates with the de Havilland Queen Bee of the late 1930s, a radio-controlled conversion of the Tiger Moth trainer. With metalwork replaced by wooden components on account of their expected short lifespans, as they were used for gunnery training.

People "in the industry" call them UAVs so I am so used to that term that drone sounds so weird.

It's been used since 1995.

http://en.m.wikipedia.org/wiki/General_Atomics_MQ-1_Predator

From that page, it seems like they didn't start sticking missiles on them until 2001, but stayed a bit hush about that for a while later. That's probably where the idea that drones in warfare is a more recent innovation comes from.

There were actually tests of Ryan Firebees firing Maverick missiles and HOBOS guided glide-bombs in 1971.

Nothing new under the Sun!

The precise variant was BGM-34 and was intended for attacking enemy air defences ( 'Wild Weasel' missions in the parlance of the day, now 'Lethal SEAD' ).

Edit: found a picture. '34A foreground, '34B rear.

https://www.google.com/search?tbs=sbi%3AAMhZZiuS8ZwZo7fS9315...

Surely the cost of military drones (and the personnel to operate them) is many magnitudes more than Amazon is willing or able to spend on delivery drones.

Surely the costs of flying a plane through a war zone with unreliable infrastructure with attached bombs is more than that of flying a plane with an ipad through suburbia with a reliable infrastructure.

Maybe not. If a drone crashes in Helmand Province, Afghanistan, there won't be much backlash. If a drone crashes in downtown San Diego, you can imagine the to-do...

You're confusing huge ridiculously expensive military drones, which are essentially small planes, with what Amazon is showing us, which are essentially small helicopters.

There's no way in the world these tiny drones can fly 10 miles each way, carry 5lbs of payload, deliver the package perfectly on the driveway/doorstep, and cost low enough to not be a financial nightmare. That's not even accounting for dealing with FAA, weather issues, safety issues.

There will need to be a few technological revolutions before it works. #1 - battery technology. Current batteries just can't handle long enough flights.

Sorry, but who are you? Who are you to stand up and call Amazon's idea unviable?

Why is their idea unviable? Because they have no proof?

Well you certainly proved how unviable the idea is with your lack of proof.

Quite frankly, when it comes to Bezos vs "ye", I have to say that the billionaire owner of one of the most successful stores in our world is SLIGHTLY more credible than "ye", random commenter who has no domain experience and no evidence for literally a single word they said.

I mean honestly, in a conversation about credibility and evidence, you thought providing neither makes for a good argument?

It's a simple question of weight ratios! A five ounce bird could not carry a one pound coconut.

Comparatively inexpensive (<$500) consumer quadcopters already have 100% payload fractions (e.g. 1lb of payload per lb of vehicle weight), so 1.5:1 isn't such a jump from there.

Sorry, but who are you?

WTF is this? I'd love to see these things flying in SF bay (windspeed:25 knts) or SoCal during the Santa Anas (30kts sustained). That doesn't mean its likely to happen with a quick and dirty solution, at least on the tech side. But who knows, maybe on the people-org side you can have a real-time delivery window that just side-steps the problem. But the appeal to "authority" here is just BS and gets in the way of a real discussion. Synchophants vs Naysayers is not a debate.

There seems to be a bee in your bonnet. The person you just berated baselessly pointed out some of the practical problems with this endeavour. He didn't say they couldn't be overcome.

He seemed to be implying it, though.

What an absolutely vacuous response. We can't possibly question the almighty Bezos once he issues a plan from on high?

Give me a break. There are a lot of technical issues to drones that should be evident to most people, no matter "who they are". I'm at a loss to understand why you think they shouldn't be able to voice that.

Why is their idea unviable? Because they have no proof?

Proof is usually a good start, yeah.

Who are you? Why are you drinking the kool-aid?

This is a very obvious PR move to get more clicks on a major shopping week and squelch a developing news story about Amazon's labor practices in the UK.

Amazon may well deliver books to me with a mini-helicopter in 2018. But they absolutely will sell more stuff due to the Prime Air attention, and will slow down the momentum on negative press that might send some holiday shoppers elsewhere. I find the idea that a billionaire owner of THE premier ecommerce business would come up with such a plan very credible.

I assumed that they would be gasoline powered. The power to weight ratio of 2-stroke engines is ridiculous, and certainly plenty to handle the demands of this sort of vtol craft.

The problem with gas and this style of aircraft is that they would need to be hybrids, or else you would have a lot of weight dedicated to very complex transmissions.

I think an all-electric fixed wing/helicopter hybridization would make more sense. VTOL is probably only strictly required at the consumer end, on the Amazon side they could do CTOL and cruise on the same wings.

>The problem with gas and this style of aircraft is that they would need to be hybrids, or else you would have a lot of weight dedicated to very complex transmissions.

I believe that on airplanes they often solve the problem of internal combustion engines being much more efficient in a narrow RPM band (with the constraint that transmissions tend to be heavy) by using a constant speed propeller[1] that spins at the optimal speed of the engine, and adjusts the blade pitch to adjust the speed of the vehicle.

I don't know if it would work on a VTOL type vehicle or not, but we've been using internal combustion engines for a while, and have reasonable work-arounds for many of their shortcomings.

[1]http://en.wikipedia.org/wiki/Constant_speed_propeller

Almost all helicopters have constant speed rotors and use the pitch of the blades to control all but 1 degree of freedom (the last is the pitch of the tail rotor blades, which directly controls the yaw in upright flight).

My understanding is that quad/octo-rotor helicopters do not have constant speed rotors because they turn about by varying the relative speeds of each rotor. They don't have any sort of swashplate.

My understanding is that this simplification is more or less the entire point of quad-rotors, since otherwise a single big rotor would be much more efficient.

Yeah, and because of that they depend on the ability to very rapidly change the speed of the props, so they need the instant torque changes of electric.

Yeah... but is scaling down the 'single big rotor' mechanisms used by real helicopters that difficult? would it be more expensive than developing a lightweight electric battery that could give you the same range?

I don't know the answers to those questions; I'm just saying, if range is the problem there are solutions that don't involve breakthroughs in battery technology.

> Yeah... but is scaling down the 'single big rotor' mechanisms used by real helicopters that difficult?

TL;DR: Yes.

TS;WM (Too Short, Want More): The top-and-tail rotor design relies on varying pitch on the individual blades of the top rotor at different points in their rotation (rather than varying speed on separate rotors) for translation, varying pitch (the baseline/average pitch of all blades throughout rotation) of the main rotor for lift, and pitch of the tail rotor for rotation.

Quad/octorotor designs get all of that by just vary rotor speed without the additional machinery needed to vary blade pitch, which makes the vastly easier to build at small scale.

The military has a much wider range of drones than you seem to be aware of. Just for a sampling of some (admittedly fixed wing): http://en.wikipedia.org/wiki/File:Group_photo_of_aerial_demo...

> There's no way in the world these tiny drones can fly 10 miles each way, carry 5lbs of payload

Today, there are prototypes that cost $2000, can carry well over 2kg (high definition cameras and such), and have a twenty miles range (that's 40 miles worth of "fuel"). As for perfect delivery… until you can miniaturize self-driving car technology, you can always have a human pilot manually landing the thing and take off. 30 seconds of human work for a small delivery, that's not too bad.

Link to the prototype.

Can't. I met the guy in person while travelling by train. He had the drone in question in his suitcase. He did tell me however he was involved with this https://www.youtube.com/watch?v=MvRTALJp8DM (Or another such effort —I met him in France.)

Ah, so you base it on what you were told by a random guy on a train.

LOL.

Random guy? Foolish mortal, you have no idea how much information he gave me. We talked for well over 2 hours. I know for a fact he has a solid technical background, and a weird twisted antenna sticking out the suitcase (about which I asked). I also fly radio controlled planes for a hobby, so I do have an idea how much these things cost, and the figures he gave me sounded reasonable.

On the other hand, to you, I am such a random guy. So don't trust me. I may not be as knowledgeable as I say, I may have misreported some key facts, I may even be lying to your face (I swear I don't, but you will never know). On the other hand…

…Do not ever presume to know how much I should know from a couple short comments. It's just not enough to tell you what I know, let alone how I think I know it.

Yes, you mention 'safety issues'. Can you imagine the liability insurance you'd need to have in order to run a business where you have autonomous flying robots with packages suspended from them flying around? Holy crap. There are so many possible points of failure.

Call me a Luddite but I will be perfectly happy to never see this become a reality.

I may not have sufficiently thought this through, but what about having them take off from and return to mobile flatbed trucks? Use them for last-mile delivery rather than all the way from the warehouse.

I doubt FAA would ever allow that.

It's a cool idea, but imagine a drone hitting a wire and crashing into a car on the opposite lane. It will kill.

It wouldn't have to be while the vehicle was in motion on public roads. Think mobile warehouse: Pull into a parking area, launch the drones, collect the drones, reload. I dunno how many dozens of packages you could deliver within a mile radius or so, but it would be a lot more than the driver could drop off on his own, which is the point. And maybe someday you don't even need a driver.

If this idea is going to work at all (and that's a big if) these things were always going to have to fly over roads and land in driveways.

Reply on news.ycombinator.com