The "Why" has to do with increasing market size for Amazon.

If they can deliver small items to you in 30 minutes with no delivery fee, all sorts of things you need in the very short term stop being a trip to the store and become a drone shipment from Amazon. This opens up HUGE markets for them.

I also think people are thinking of the distribution network a bit wrong. They can keep a large number of items in a small space, particularly if the warehouse becomes sort of like those old magnetic tape storage robots you used to see. You could condense an entire Wallgreen's (almost every item of which is drone-able) down to a large room. So the practicality of having a number of them in a city becomes more and more viable. Of course you'd have software optimizing for the most common items on a per-distribution center basis (I bet Amazon already does this on a macro scale).

Add in long-haul robotic trucks feeding the distribution centers and you can see where this is going.

Yeah they can make use of the model employed by tesco in South Korea[1]. Instead of trips to the grocery store, you can just tell amazon to deliver to your place. Ran out of batteries for your remote? Amazon! Heck, Amazon Book store would benefit LOADS from this. So yes, not only does this open up new markets, it also improves their existing businesses as well. No one likes to wait. And they could cash in on that! (Not to mention it would be pretty cool receiving stuff by drones) :P

[1] http://gajitz.com/worlds-first-virtual-grocery-store-now-ope...

I can see an Amazon Prime subscription coming with some type of landing pad dongle. You lay out your mat & dongle somewhere, put in an address and the drone flies to the address, listens for the signal from the dongle and drops the package on the 'X' if the details match. My bank currently gives me a two-factor dongle for secure logins, same type of approach for authenticating delivery and indicating a safe landing area.

There is a lot of really interesting problems to solve to turn this into a commercial service. Would be very fun to work on.

The last mile, the landing, might one day be a combination of a drone and a vehicle. It can fly but land and drive up to prevent injury, and use driveable tech as well as drone tech. It would advance both fields and maybe get us to flying cars one day, of course we'll probably never fly them ourselves, just computers. Autonomous drones and delivery vehicles would be awesome.

This is a very nice mod of the flying drone delivery idea: flying in the long range, over safe zones like highways, and then scooting down the last mile in densely populated regions.

>The FAA situation. Drones are going to have _very_ restricted airspace especially in populated metro areas, if the FAA grants them these privileges at all. If you look at Los Angeles/SF/NYC/DC/Boston, you see airspace that is so overcrowded that I just don't see how drones are going to be zipping around here.

How do you figure? Do you realize how big the sky is? For every square foot of land there are tons of layers of airspace above it, and it's almost completely unused except for a few tall buildings and a relatively small number of flying vehicles (compared to what we manage to fit on the ground.)

Drones are fairly maneuverable and could be made with decent collision avoidance and sensors.

In addition to sensor-based collision avoidance you would probably want some means of communication between the drones so they can choose non-colliding flight paths in advance.

In the altitude these drones will be flying most of the regular traffic probably is composed of helicopters. For those Amazon could easily provide some means of requesting a temporary drone no-flight zone.

Ummm.... I couldn't see that video (restrictions on work internet) but AFAIK such small drones can fly up to a certain altitude. Sure you have an insanely large amount of sky up their but can these machines access that?

Imagine how many of these could fit on a single flat road. Then imagine stacking dozens of roads on top of that. Then imagine the roads are several times wider because there isn't any property or buildings in the way up there (with the exception of a few tall buildings in big cities.)

It will be a long, long, long time before we run out of airspace. It isn't the limiting resource at all.

I never doubted that we dont have enough of airspace, my question was can these lightweight machines make proper use of all that open space? how high can they fly? Because i am assuming these are relatively low-altitude aircrafts.

With respect to (1), why would delivery drones fly at high altitudes? I suspect that the only overlap in the flight envelopes of delivery drones and of manned fixed-wing aircraft will be riiiight around the airport, and during horrible plane crashes.

(Manned helicopters are another matter, but how many of those are there really? Even in the big cities.)

I think there's a significant long tail for products that may be eligible for 30-minute delivery. Pretty much the only time I go to a real store to buy something non-perishable is if I'm on a time constraint and can't wait for Amazon to ship the product.

If I need a cable or other electronic peripheral that isn't carried by big box stores, there's exactly one vendor in my metropolitan area that is likely to have it in stock, and they'll unabashedly charge 5x-10x what you'd pay to buy it on Amazon if you had 2-5 days to wait for the package's arrival. I'm sure there are many markets like that right now.

The potential this has to decimate many traditional retailers is super intriguing. What are we going to do with all of that extra space?

This has been coming, and there's also so much room for gains from it. Eli Dourado covered this nicely, I think http://www.wired.com/opinion/2013/04/then-internet-now-airsp...

Personally I hope the commercial use of drones become evident and widespread and legislation lags behind it. The last thing we would want is strict military-only licenses for drones out of the fear someone will deliver some explosives to a politician with it.

It is pretty awesome, indeed. I'd say your issues aren't that big, though:

1. FAA is the toughest challenge, true that; 2. A good idea would be to use the roofs for delivery; 3. To add to that, people will buy a membership just for the convenience of getting a lot of small stuff in less than an hour without leaving the house/office. Hard drives, screw drivers, nuts and bolts, stuff for BBQs, etc.; 4. Stealing them will likely end with them being tracked by GPS, found and fined, people will learn not to do it soon enough :); 5/6/7. I think the failure rate will actually be extremely low - quadcopters are very resilient, and we've already seen how amazingly well they can handle the weather, bumping into buildings and flying with three rotors - obviously, Amazon will have to go with the best hardware design;

I think this could be big right away - as I said, many people want immediate delivery and with a subscription model, Amazon won't even have to ship all that often to start making money off of this.

Most apartment buildings don't have accessible roofs, at least where I live.

But if that were possible, it'd be the safest altitude given they manage to avoid wires and antennas.

If they make it to ground level, I think there'd be plenty of people who would try to catch these things.

I don't know about the US, but in the UK there are 'sky lanes' that helicopter pilots must follow.

And in all honesty, this would benefit Amazon, because then they don't require drone pilots. Just programme the 'lanes', rudimentary GPS and proximity sensors and keep the speed down so that collision-avoidance can be activated with sufficient time.

Birds.

I can imagine FAA approval subject to a register of drone flightpaths and locations, but if something takes a knock from a bird (or malicious interception, as you suggested), I assume that Amazon would schedule a manual pick-up and replacement, wearing the cost in the process.

"You'll have kids/idiots trying to steal and shoot these things down." agree. the fact people can steal it is an issue.