Hacker News new | comments | ask | show | jobs | submit login
Amazon Shows Off New Prime Air Drone with Hybrid Design (amazon.com)
281 points by threecoins on Nov 29, 2015 | hide | past | web | favorite | 168 comments

Here's what I think is more likely:

A small regional airport is taken over by Amazon. They build hangars and warehouses onsite.

It's ramped up to land and fly a big turbine-powered mothership every five minutes. The warehouse dispatches packages to stacks of delivery drones like the one pictured in the video. The mothership loads the delivery drones in order. The motherships take off regularly, and then as the mothership flies near the target swath, the drones are released one by one and glide to the target zone efficiently before doing a vertical landing.

Empty, they return autonomously to a small recovery facility in a vacant lot somewhere within ~5km. They're packed into a pickup or a semi and returned to the airfield in stacks.



Because VTOL flight is not easy to do at long ranges; Batteries generally limit that to 10-30 minutes since vehicle mass fraction is large. Because very small planes are only a small multiple (2-5x) more efficient than VTOL flight. Because high speeds are something vehicles like this are not great at.

A drone like this doesn't have a great straight-line range when you sling a heavy load under it; Amazon is not going to erect sizable warehouses within ~5km of every plausible delivery. Their stock is too large, and too sparsely ordered.


Or, even more likely?

Replace the flying mothership with a box truck dispenser. The driver stops at his target neighborhood, sends the drones out, they deliver autonomously, takes the drones back in 10 minutes later, and returns to regional warehouse in the next 20-50 minutes of driving, for 30 to 60 minute delivery.

Amazon gets to deliver to a neighborhood rather than to a house.

I believe your scenarios (especially the "truck dispenser") address the market Amazon has in mind: the suburbia.

If you notice, both video examples show the drone delivering at a house. It seems obvious that an urban setting is difficult to address. And actually delivery costs are higher in the suburbs, due to population spread, than in the cities.

The interesting thing is how do UPS/FedEx/USPS feel about this, considering that Amazon is aiming to "evolve" their business model? We should soon hear about FedEx/UPS trying to do drone delivery. But this is not only logistics but also a lot of other software and I think Amazon might be better suited for it.

I was under the impression that UPS/FedEx are going to merge to form UP/Ex, which will utilize railgun launchers and aerodynamically-self-steering packaging to deliver anywhere in a 100-mile radius from a single facility?

Using the autodidactic deep-learning A.I. that "corrects" it's targeting over time through an IoT police scanner.

Is this a reference to Rainbows End or am I just imagining things?

USPS won't care, their (legally protected) business is shipping letters.

I always thought USPS would be a perfect partner for Amazon since their post offices would be ideal takeoff points for the drones. They are everywhere and they already deal with receiving packages.

USPS was worried enough to agree to start delivering packages for Amazon on Sundays, which must've required some interesting negotiations with their unions.

Sunday delivery workers are paid just above minimum wage and are non-union.

Delivering letters costs the USPS more than they charge customers; they make money on package postage and use that to offset the losses they incur delivering mail.

Based on a comment in this thread which suggested that USPS would be a prime (no pun intended :) ) candidate for a partnership, especially due to the existing real-estate infrastructure. If we think about it: USPS can very easily deliver letters/magazines using Amazon's service.

What if Amazon soon will develop the AWS-of-postal-delivery infrastructure?!

'The interesting thing is how do UPS/FedEx/USPS feel about this, considering that Amazon is aiming to "evolve" their business model?'

UPS/FedEx/USPS can feel whatever they want but there's not a darn thing they can do about it.

I'd guess they'll be elated. Delivering packages in the far reaches of suburbia (and beyond) is probably costing them money, as it takes a lot of truck&driver time to deliver a single package. If we assume they maintain their hold on urban deliveries and let amazon take care of the suburbia ones; their margin will go up.

Also, I'll point out that in my area of suburbia, UPS/Fedex sometimes deliver to USPS and let USPS do the last mile. it's not hard to imagine they'd deliver to an amazon base station where robots & drones would take over.

What Fedex does in my neighborhood I haven't seen prior.. they recently brought a gas-powered golf cart and the driver uses his truck as the "warehouse" and makes rounds all day.

UPS has been experimenting with drones for a while.

I agree with the last one, but ground delivery robots would make way more sense. I don't know about 30-60 minutes, however.

The soccer shoe emergency can't really be that big of a market. Emergency medicine by drone delivery I get, but the reason people go shopping nowadays is either grocery or an excuse to get out of the house.

Drones are never going to eliminate the excuse to get out of the house factor.

Today's soccer shoe emergency is tomorrow's normal turnaround. Imagine trying to explain to someone living in the 90s that they just have to watch this one TV show right now.

Surely they could just walk down to their local video rental store, not that many people really need to watch something right now!

But once something like BitTorrent & Netflix happens it becomes the new expectation. I think something like drone delivery or delivery via autonomous vehicles is going to similarly change the market for the delivery of physical goods.

To further elaborate your point, I think it is also about reducing friction. People will probably still go grocery shopping but if you were fairly confident that something you are not sure if you need is available both at Wally's world and at Amazon.com for about the same price, would you rather buy it anyways during your grocery run or put it off until you know you will need it?

I don't want to think whether I will need something near the end of the week. We throw big words like nothing but I concede this is one of the hard problems where the idea of express shipping (drones, land robots, or even delivery humans) in conjunction with large distribution networks could have profound impact on the way we shop.

While the distribution warehouse infrastructure of Walmart is probably nothing to laugh at, I imagine a large portion of the infrastructure is not centrally owned and managed by Walmart. Just like the famous "the network is the computer" motto, I bet Amazon thinks the warehouse infrastructure is the retail store.

(To go a little off-topic, the physical Amazon.com book shop sounds strange to me. I don't get it.)

This will sound a bit off kilter, but except for truly emergent situations (injury, medication, etc.) it's good to have some "friction" otherwise we could end up with a population demanding immediate satisfaction and perfection from everything, whereas many things will remain imperfect and unrealized. Why can't my government change right now the way I want it! Why are they serving someone else in need and not me!

I have a fear we're on the way to creating a whole class of people who are totally dependent on the system for everything. There will always be people who are well rounded and know the relationships between things (i.e. know where the sausage comes from)

Why shouldn't we demand immediate perfection? Why should we have to choose between serving someone else and serving me? Why shouldn't governments immediately change?

All the problems with having these things today are implementation problems not fundamental issues.

We want a delay between what we think we want and what we get from government, else we get majority rules democracy which is not far removed from mob-mentality government.

I think you're stretching a little to get from faster merchandise delivery to mob-rule government.

Or, your objections would have been valid against the railroad and the telegraph as well (both of those reduced friction tremendously).

I get the gist of what you're saying, but I don't agree that artificially limiting technological innovation, or improvements in efficiencies is beneficial at all.

I think railroads reduced times to human times (from weeks/months to hours/days --but it wasn't a snap of the finger kind of response.

The objection I have is more with the psychology rather than the physicality of it. It makes things more disposable. You lost a sneaker, we get it replaced in minutes. It's allowing people to gorge on consumerism. It also puts the local shops at a disadvantage. But, mostly, it removes the barrier to thinking about buying more. You don't have to think about waiting till tomorrow, or two days from now, or having to walk twenty minutes, or get out and take transit, or jump into a car to go satiate the impulse. No, all you need is your desire (and buying power) and it's fulfilled.

Fair point. This just happened to me yesterday. I needed a 3.5mm to RCA cable to connect my TV to the receiver[1]. The one I had wasn't long enough, so I found one on Amazon and ordered it for the next day.

By the time it was delivered, I had already found my long-lost extension cable and didn't need it anymore.

Waste of money and—to your point—if I didn't have that option I probably would have searched a little more around the house before resorting to buying a new one.

Ultimately I still want to see these things advance, though. Whatever mechanism that would set lower bounds on delivery times would be worse than the problem it would be intended to prevent.

[1] Never mind why the TV only outputs audio on a 3.5mm jack, kinda cheap.

That ship sailed with plumbing and the electric grid.

One word (well, acronym): DIY. Middle-class suburban dad is trying to fix the newly-bought house, but he's constantly running out of the right size of screws and/or stuff he didn't even know he needed. At the moment that scenario results in a a lot of trips to the hardware store, but cheap 60-minute deliveries would change it for the better.

Neonatal supplies, cleaning supplies, you name it -- the sort of thing you really don't want to shop for, it's bulky and you never know if you're getting the right brand: if you could order it from a bookmark, knowing you'll have it in 60 minutes, then the actual shopping experience would become much more fun, because you can actually pay attention only to the funky meals you'll cook and the new stuff you can try.

And of course, geeks would jump on this like crazy -- 60-minute deliveries for new hard disks, mice, games and any other soho/gaming equipment? I would sign in blood ink right now.

I have my doubts the drone model can scale (for example, in Manchester, UK today we had 80mph gale winds...), but it's an awesome idea that would definitely change retail.

I drove to the hardware store three times during the course of a single plumbing project yesterday. Each time, I bought everything I thought I would need to finish the rest of the project! I'm not even in the suburbs, I'm right in the middle of Seattle. I definitely would have paid extra for drone delivery instead of making that third trip, because I was really tired of driving across town at that point.

To be fair, in the long run the DIY use case is more likely to be served via personal fabrication. For small finicky bits like screws, being able to print-on-demand is going to be quicker and easier than waiting for a drone.

lol delivering a physical game

A 20GB game (which is far from unusual now) in 30 minutes = 89mbit/sec, which is eight times the US national average broadband speed.

Ever tried to download a boardgame / tabletop? It's hard :)

I suspect this will be limited to urban areas, like all of Amazon's premium delivery options. 30 minutes of flight time at 30 mph means you could cover the entire city of Houston (the largest city by area in the U.S.) with about 4 warehouses. I think you're much better off trying to increase the range of your drones than using carriers because your coverage area is quadratic with your drones' range but linear with your number of carriers.

Drones have got to get back in addition to getting to the target, and you only get 30 minutes if most/all of your payload is dedicated to battery. The drone has a fairly short range while fully loaded, and a lot longer range while empty.

86% of Amazon's deliveries are under 5lbs AND a glider style drone has way longer flight times than a traditional quad/octocopter.

5lbs is a large number in this context.

What we need is a sort of tcp/ip network for drones where they understand where charging stations are along the way, and other brands delivery drones would eventually be able to participate in the network as well which would spur growth. I'm good for a few charging / delivery stations in my neighborhood

> coverage area is quadratic with your drones' range but linear with your number of carriers.

How running cost scales with drones' range? If it's also quadratic or worse then there is no reason to increase it.

Or even better, Amazon could build housing around their warehouses and move their shoppers into them. No flying motherships. No drones. No UPS necessary. We could call them Amazon Prime Cities. 100 bucks a year, housing included.

As a one time hooliganish teenager, I wonder if these will become targets for slingshots (or more sophisticated surface-to-air weapons).

Not just hooliganism: there is real booty to collect.

I suppose the flying altitude should be outside te air gun range, with descent right naivete landing spot.

You can also collect booty from parked UPS and FedEx trucks.

In San Francisco, at least, the trucks' cargo areas are locked and reasonably well-secured when the vehicle is not attended, to reduce the opportunity for exactly this crime. I'm sure there are areas where sniping a few boxes from a truck as the driver is otherwise occupied is not so difficult, but that's not something one could expect to get away with repeatedly.

There's a reason the drone is made to land going directly down, vertically, once it reaches the destination.

If drones do get shot down, I'm sure Amazon will eventually deploy escort drones that sense where the shots come from. lol...

Hitting something small going 30MPH at 400 feet seems hard even with most guns. Adding some cameras you could probably track down who is attacking easily enough to discourage most people.


Drone is a smaller than amazon's but it's a good demonstration

When Amazon first announced their intent to use drones, one of my friends started using a phrase: "Skeet shooting, with prizes!"

From the page: "Our vehicles will be built with[...] sophisticated “sense and avoid” technology"

I can see an arms race between Amazon and would-be air-pirates; I honestly don't know what the state of the art in radar technology is, but if it can be drone-mounted, then you'll need something seriously fast to hit an intelligently-dodging vehicle.

Google showcased a radar-on-a-chip during the last IO; it looks exactly like something that could be fit onto a drone and provide short-range monitoring for incoming objects. The issue would be drone's reaction time though, especially when carrying heavy load.

As for hitting an intelligently-dodging vehicle - ever tried to throw a stick at a pigeon? I obviously don't recommend doing that, though I have some personal experience from the time when I was 9 years old, and I can tell you that suckers are damn good at avoiding projectiles. I doubt drones will reach that level of performance within next year or three, so for now, if you have something you could hit a pigeon with, you'll definitely be able to take a drone down with it.

Did you shoot at cars with a slingshot?

I assume this is for bigger cities though. In Chicago you can get to over 2 million people within 6 miles as the crow flies. If you can dodge powerless and buildings it seems reasonable...

I wonder how many of those 2 million people have an appropriate size space for this video's "landing zone in the back garden"?

Where I live is probably high enough density to make this idea work, but there's not enough landing space I think: https://www.google.com.au/maps/@-33.9038146,151.1761515,327m...

The drone could hover over the sidewalk in front of a building, you could signal it with your smartphone similar to how the sign is used in the first video. The drone could then drop the package or it could have a mechanism that allows you to pull the package from the drone while it's in the air above you.

Heh - have you ever been close to a quadcopter capable of carrying a 2lb payload for 20+ minutes?

It's pretty much 4 carbon fibre lawnmower blades thrashing around in the air. I'm not getting anywhere near close enough to remove a payload from one while it's attempting to keep a stationary hover in an area of poor gps reception.

I'm reasonably careless with my little quads (~350g with 5 inch plastic props spun by 220W motors), but my medium sized one (~800g) gets treated with much more respect. I've got a friend with a 5+kg Octocopter - it's got 2+kW motors spinning 18 inch carbon fibre blades, I don't get within 4 or 5 meters of that when it's powered up and armed (and I'm extremely wary of it when it's got power connected but isn't "armed", because that's really just trusting Chinese electronics and open source software to not chop my fingers off...)

I admittedly haven't been that close to one but we do trust or are seriously considering trusting software to control heavy machinery like cars. Is is truly so out of the question that we might be able to make them safe enough for that?

It might require more advanced designs, far better tested and well regulated software but being able to get that close to a drone would enable a lot more use cases for them. I think many people are going to invest quite a lot of money into making this possible, so I think we're going to see something like this eventually, even if it takes decades.

You can mitigate this a lot with more props of shorter diameter moving slower, and grilles to serve as prop cages, at the expense of somewhat lower flight efficiency.

Good point, but Amazon's drones are probably going to be propelled by ducted fans for just this reason.

I can imagine flats in SF (for example) being advertised as having a prime air landing spot on the roof

It's worth noting that Amazon already has warehouses in major urban centers for Prime Now: Nashville, Richmond, San Diego, Portland, Los Angeles, New York City, San Francisco.

It seems plausible that by the time (i.e. in a few years) Amazon gets regulatory support for its delivery drones in a suburban community that it'll have saturated urban delivery markets in the US and will be ready to build warehouses in the suburbs.

Internationally too. Three major ones in London so far, used to support Prime Now as well as same day delivery for a large proportion of "normal" Prime.

>Because very small planes are only a small multiple (2-5x) more efficient than VTOL flight.

Seems like a longer wingspan is the simple answer, no?

its a delicate and complex balancing act of weight, lift, drag, and thrust of both the airframe and any rotors, and the power output and energy stored aboard the craft.

longer wingspan means greater lift, allowing more payload, but also means more drag, and the additional payload and longer wings also increase the mass of the craft- leading to greater power requirements, which means more powerful motors, larger rotors, and larger batteries, which are also heavier.

The point is, drone/aircraft design and logistics like this cannot be solved with a simple answer like "longer wingspan".

How much longer? What about all the cascading effects on all the other aspects of the craft? This all needs to be addressed and optimized to be implemented properly.

I believe the truck dispenser is the most likely scenario. I bet $1 on it.

Why would the driver need to stop?

Except there will be no driver.

Cue Amazon's PR firm trying to win 'cyber-monday'.

Whether or not Amazon ever delivers packages with drones, the ROI on the drone R&D is massive simply based on how much attention it draws each year on cyber monday.

Amazon is such an interesting company to watch. They do really dumb and goofy stuff, all the time, in very visible ways, but under the radar they've got the largest cloud hosting business in the world chugging along devouring everything in its path that most people don't even know about. I mean, none of my non-technical friends know AWS exists or what it is, and few realize Amazon is one of the most powerful and influential tech companies in the world.

This drone stuff even feels like a distraction so that nobody notices what Amazon is actually up to. It's currently so far from useful, but it does generate a lot of PR for Amazon, and the timing is not at all subtle.

It's almost Christmas so it means another video from Amazon on their futuristic drone tech. Sigh.

Most news stations in every country will probably tag this to the end of their broadcasts thus meaning millions saved in advertising for Amazon.

I like the way they make it sound like the only thing stopping drone delivery is government regulations:

> We will deploy when and where we have the regulatory support needed to safely realize our vision.

While I find the technology impressive, it makes me wonder if "Black Friday" is going to look like the Blitz, a low hum and looking up, wave after wave of aircraft coming over the city looking for the right place to drop their payload ...

I like the placards placed in "safe landing zones". Simple solution to hard problem.

I remember when Amazon first announced the idea two years ago there was a lot of brainstorming in the hn comments and that idea was pretty popular. No idea if it influenced Amazon but would be cool if so.

*found the thread https://news.ycombinator.com/item?id=6833223

"This is just one idea. Perhaps Amazon provides you with a beacon for the drone, or you get a big QR code for the drone to target your house & they don't need pilots at all. Who knows!" -owenmarshall

And mess with your neighbor's life by placing your own landing placards in your fenced-off backyard.

I assume that could be prevented by using QR codes or something.

This fails if:

1) If you can see over the fence...

2) If it shows up in Google Maps ;)

#2 might be implausible now, but if people get really accustomed to this, I guarantee that some handymen are going to tile/spraypaint large, permanent drop points in their yards.

Yeah, wouldn't be better to have one that has encoded QR code with the package information?

How does this "mess with your neighbor's life"? I wouldn't see this being a huge concern unless you happened to place 4 or 5 orders a day.

I think the poster was implying that you trick the drone into delivering your neighbor's package into your backyard by putting the card out even when you're not expecting a delivery.

GPS is way more than accurate enough to discriminate between your backyard and you neighbours. Even with poor satellite geometry I rarely see more than 1 or 2 meter CEP when I'm up above the tree/building line.

I'm betting that by the time this is real (and assuming the shown implementation is what actually goes live), everyone will have their own personal QR code so there will be no mistakes.

Take a pic of your neighbour's QR, print a copy, profit.

Wait 40 minutes, don't get package, email support, they review footage/delivery data, see drone going to your neighbor, press charges for petty theft.

Did I say anything about keeping or even opening the packages? If the drone is using some sort of convnet or HOG to detect the landing pad, I can only imagine hilarity would ensue if its visual field were saturated with multiple copies of the target image, a sort of visual DDoS if you will...

Yeah. And even for apartments, you could use window support brackets (like those used to hang AC units, but removable).

For a while I have thought that a cool way of doing this would be if the drones could fly close to high voltage power lines and use induction to power it and charge the battery. Then when it gets close to the destination it could peel away from the high tension lines and use battery for to cover the last mile. That way you could have the warehouse in some industrial area or whatever. They could make a deal with the power companies and those lines could be like the arterial lines for the drones. I have Googled around and this is not a new idea but I have not seen anyone try it or test it, at least not from the searches I have done. I believe it is feasible as those lines are 500KV and as a drone moves past it, there is a significant EM field around it and an inductor lowered down 20 feet or so from the drone could pick up enough power while maintaining a safe distance. Has anyone else heard of such a thing being tried anywhere ?

I once worked in a building that had one of those power lines running next to it. Was curious if there was a measurable effect on a car parked under it in the parking lot. A multimeter found 600 volts between the car and ground.

Wouldn't that pose problems for the navigation equipment?

Depends on how the IMU and GPS are designed. It could pose a problem for equipment that isn't purposed-designed to ignore that frequency. However, textbook solutions for that problem exist. 60 Hz is well outside the vehicle's closed-loop control bandwidth, so just ensuring that there isn't an aliasing problem with that (or the rectified 120 Hz) would be sufficient.

My major takeaway is that they are trying to push as close to instant delivery as possible (30 minutes seems like pretty remarkable goal). Potentially taking out the delivery service middleman is also good. I have my doubts (regulation, security concerns, social doubt a la Google glass) but overall it seems like a good long term strategy to try this.

If they ever come close to hitting the 30 minute delivery it also opens up the entire food delivery industry as a potential Amazon extension. Buying one of these "premade meal delivery" startups could also be an option.

> If they ever come close to hitting the 30 minute delivery it also opens up the entire food delivery industry as a potential Amazon extension. Buying one of these "premade meal delivery" startups could also be an option.

This is my thinking as well. I'm actually surprised they haven't started a small project in a specific area where they're delivering hot meals. It feels like they want to, eventually, sell everything possible to people.

I would actually be surprised if they don't, eventually, buy several of these start-ups around prepared food delivery.

Just imagine ordering your hot pizza delivery from Amazon instead of Domino's for example.

Amazon already does meal delivery in Seattle.


Bingo on the food. Literally none of these food delivery startups will be able to compete with the speed, price, cash, and logistics infrastructure Amazon will be able to bring to the table for that area.

Delivery people, like cab/Uber drivers are service industries chugging along on borrowed time until their jobs can be fully automated.

Love that they got Jeremy Clarkson from Top Gear fame to do the video. I'm guessing he likely had a hand in writing the copy as well.

Amazon signed Jeremy to do his new show on Amazon Prime, so I'm expecting to see more Jeremy/Amazon promotion going on.

Violent jerks have no place in a civilized society.

I think there's gonna be alot of teenage pranks and destruction. People setting the landing spot in the water or people stealing them with nets. Not sure if much recourse unless there camera footage or gps locator.

This is the standard fear response to new technologies. Uber: how can you get in a car with a stranger? Airbnb: Even Paul Graham said "are you nuts?". Self driving cars: but what if people take advantage of it's collision avoidance abilities to walk in front? etc. etc. The fact is, most people are not malicious, life goes on anyway, and the risk of vandalism is quantifiable and included in the cost of doing business.

And in this case vandals can easily be prosecuted, since a drone is already hooked up with remote control, GPS tracking, cameras and monitoring.

I find a lot of the whole drone delivery thing pretty silly outside of some specific scenarios. Nonetheless, there is an equally silly "we must guarantee 100% reliability and security" thought." UPS does fine. They have no trouble leaving most packages at my door. They probably don't live the same thing at other doors. (Though I don't know their heuristics for such things.)

OTOH, I feel like if people today proposed that trucks leave unattended packages outside people's door, it would be dismissed as unworkable. But it happens constantly with minimal problems. There aren't as many bad people in the world as some people think.

I've got a 100% theft rate for unattended packages in the past 90 days.

Thank you captain anecdote.

The OP must not live in a city. Package theft is a cottage industry.

The drones are almost certainly going to have cameras (for object avoidance) and GPS on-board in order to satisfy regulators that Amazon knows where its drones are and that they're safe. Consider Amazon's own airspace proposals linked in the FAQ of the article.

No way. There will definitely be camera footage- multiple cameras on the drone getting multiple angles- and more importantly, from other drones.

Along with several GPS devices on each drone, each with redundant power supplies.

I think someone previously made the analogy of a lot of UPS trucks getting stolen. It could happen, but prosecution is probably a sufficient penalty to prevent it from being a serious issue.

Also easily fixed if the FCC makes messing with unmanned aircraft a felony.

Sort of how teenagers occasionally destroy people's mailboxes today? There are risks to everything and we got over it.

Yeah, I have a feeling there's going to be a lot of drone hunting going on, until Amazon figures out a way to secure these things.

I think there's going to be a few very high profile prosecutions of drone vandals.

Shit, I thought this was a marketing gimmick. Looks like they might actually be delivering. And it looks quite awesome to be honest.

And development in Israel? Wow!

I don't get it. What so special about development in Israel?

That means they'll have onboard countermeasures for surface-to-air missiles.

It still could be. They did it right after thanksgiving last time too.

Yup, right before Cyber Monday.

Every neighborhood should have a delivery nexus where packages are to be picked up / dropped off! That'd make more sense than everyone needing a dedicated Amazon landing pad in their backyard.

You could also have fresh baked goods and produce economically delivered daily like they did back in the cutting edge of the 1800s.

Like a post office?

Yes, except the idea is to have one within walking distance for most people, and it could possibly be automated and unstaffed.

I wonder how can delivery with drones be more economically effective than with cars and drivers, with "distances of 10 miles or more"? Which, if I'm not mistaken, still somewhat exceeds capabilities of today's drones.

People are expensive.

They're not that expensive and are very flexible dealing with the very analog common-sense-requiring situation that is dropping off a package.

They are much, much more expensive then drones. Minimum wage, benefits, delivery vehicle maintenance, insurance, etc etc etc.

Yes, people are more expensive. In that they're paid wages. Others things you list are common to automated platforms. But people are also good in dealing with imprecise situations that are common with package delivery. I actually believe Amazon is seriously interested in pursuing this. I also believe it's largely stupid.

Do you know why medic-vac helicopters exist? Not necessarily because they move faster than an ambulance, but because they can travel as the crow flies. Even if a drone and a human delivery driver cost the same, a human could never come close to the delivery speed of a drone simply because they can fly over stop signs, stop lights, inefficient road design and get to the home the fastest. And for Amazon, that is the value, getting the package to the customer as quick as possible.

Yes and medic-vac helicopters also exist because they're responding to genuine emergencies which justifies both the expense and any externalities (e.g. noise and marginal danger) associated with their flying over houses etc. to save a few minutes. Sending flocks of flying vehicles over houses to deliver cables or pizzas because people need them RIGHT NOW is not the same.

I do see scenarios where drones make sense but I have a lot of trouble envisioning people being fine with the skies of suburbia being filled with drones all the time.

They are not very loud; they have a smaller impact than a UPS truck or a pizza delivery guy racing against time. They will get lots of vehicles off the road. I see folks embracing drone delivery as a solution!

Oh fill the sky with drones, no worries there.

If they start crashing on my head during a picnic, I would be a little annoyed...

"Jeff Bezos beats DoD at it's own game" -Business Insider

Drone delivery without explosions.

I'm really curious how this beats a amazon owned delivery truck? Serious question.

Economically and at scale, it seems like both the truck and drone are actually pretty cheap.

Speed of delivery and at scale, the only way the drone is faster is if there is a warehouse in each city and your item is stocked. Otherwise, the bottleneck in delivery time is the long distance shipping.

Feasibility and at scale, the truck seems way less risky and more adaptable to bad conditions.

Often going 5 miles in a city can take 30+ minutes. So using drones makes this easier. Further per package delivery costs are probably lower. At day 10,000$ per drone, an average deliver 8 packages a day, for 3 years. That's 1.25$ per package.

To be economical, the truck would have to be relatively full, which means dozens of dropoffs, which means the average delivery time is measured in hours, not minutes. With a whole bunch of drones, you can do all those deliveries in parallel.

I don't think this will be everywhere, just urban -- at least a t first. My guess is they'll also use PrimeNow (1, 2, 4 hour delivery by car/truck from a localized warehouse -- happening NOW in my city) as a canary in a coal mine to test a given market's suitability to immediate delivery (in terms of demand, regulatory env, probably tons of other factors).

With automated drones and Amazon brick and mortar stores the idea of large warehouses could be challenged. They could shift packages in an automated fashion to areas with more probable demand. Drones are crazy cheap as well. I can imagine hundreds of these in the sky at once in Chicago if not thousands.

>I can imagine hundreds of these in the sky at once in Chicago if not thousands.

Which sounds like an utterly horrifying scenario. It's not so much that I can't imagine this sort of delivery not working in some scenarios as that I have a lot of trouble imagining it working at scale.

people in the 60's thought we would have flying cars in the 21st century.

in reality we'll never need to go anywhere anymore, not for work, not for meetings, not for a movie, and not for groceries. what's a city for anyway? Oh, nightclubs I guess...

A couple days ago, I needed a new cordless landline phone and wanted to try out Prime Now. It was impressive that I could track my courier's location and have it arrive within half an hour - only problem was that it was the wrong model. I've ordered hundreds of things from Amazon and can't recall ever receiving an incorrect item. To make matters worse, the CS rep I spoke to said they couldn't send couriers to do returns or exchanges: I had to UPS it 2,000 miles to Kentucky. Winning the ecommerce delivery race won't mean anything unless Amazon can maintain its high accuracy rate and positive user experience.

It would seem to be a better option to try and drop a package by parachute rather than have to land the drone, release and takeoff. If the package is only 5 lbs. or less and has a parachute attached that seems much safer (for both the drone and for people) than landing and taking off assuming a predetermined landing spot is located.

My other issue with this is given changing weather conditions and the amount of wind or rain a drone can handle (energy efficiently that is) how many days per year can this service be offered vs. how many customers there actually are that need this type of delivery. (Note the example that they give ..)

> by parachute


> wind or rain

Obviously parachute delivery is more sensitive to weather conditions. Parachutes would need active guidance to be at all practical. If the parachutes and guidance systems aren't recoverable (collected from your mailbox, maybe?) that'll add to delivery costs.

And on the day before "Cyber Monday," what an opportune time for Amazon to dominate headline for a day. Such luck.

Does anyone else think that underground tunnels are a superior way to solve the last few miles problem? Tunnels about 1 foot in diameter with little wheeled vehicles pushing around payloads of takeaway food, groceries , parcels at reasonable speed with control computers determining routing and stopping collisions. A functioning network of this type would be more valuable than power or cable utility networks, and horizontal tunnel drilling tech means that you dont have to dig up huge swathes of road to put tunnels in.

Creating tunnels would be several orders of magnitude more expensive than creating the technology for drone delivery, lobbying for it's adoption across various municipalities and so on.

More likely than not, these drones are for the rural last mile problem, where drivers aren't worth the cost. In that case tunnels wouldn't be worth the cost.

It's most likely the last 20 miles problem where delivery costs are a problem. I live in an exurban area but I imagine UPS economics works fine here. If I live in the back of beyond I imagine it's hard to make any delivery system work without subsidies.

New York City tried something like this with pneumatic tubes for delivering mail.


Seattle's current tunnel project isn't going all that well...

I wonder what the economy of dealing with stuck packages, flooding, rodents etc. would be...

Is it just me, or is Jeremy Clarkson's voice off by a few milliseconds in the video?

Ya, I noticed that as well. Very irritating. The service looks awesome if they can, uh, deliver though.

Thinking about it a lot, and I think the simpler way to solve the same problem would be shipping container sized package 'vending machines'. These could be distributed anywhere a couple parking spaces worth of land could be rented.

You could either pick up your items yourself, or pickups could be handled via an "uber for packages" type model.

The advantage of a system like that would be how quickly it could change based on demand, and it would work well inside all the current shipping infrastructure.

The trick would be stocking the right items in the right locations, but if anyone has the data to be able to do that it's Amazon.

It's called Amazon Locker, and it's existed for years.

Amazon locker just swaps the last step in the delivery to a central location. I'm talking about micro-warehouses that could easily and quickly distributed.

I have trouble understanding where the package goes in the drones (which I guess is good for deterring theft). Also, they don't really provide any scale, but the drones appear to be pretty huge.

For scale reference: towards the end of the first video [1], they show the drone landing on a pad next to a fence, and a few seconds later, a woman picks up the box and pad.

I'd say it's about 6 feet on a side, and maybe 4 feet tall.

[1] https://youtu.be/MXo_d6tNWuY?t=107

The drone looks like it's roughly 2.5m wide and 2m long

While time will tell if we ever actually see this drones in action, imagine how cool it must be for the people working on this project. They get to go to work each day playing with some really fancy model airplanes. Plus, this is Amazon pushing yet another aspect of technology forward. They may overwork their employees, but cool things are coming out of those offices.

Probably won't be able to deliver in bad weather. Small aircraft aviation is always subject to the whims of the weather.

This made me think about self driving cars and whether its possible to make a self driving delivery van or truck - able to handle bulkier / heavier loads and all weather conditions.

The big question is: How loud are the drones? Every quad-rotor I've been around is too loud for the average upscale neighborhood that these will be delivering to.

so, assuming this works: Won't they get a lot of problems with wind / bad weather regions? I guess they won't be happy if they can do this only half of a year, and the other half the drones are sitting around.

So, if this takes off (phun intended) will Amazon try weather control next?

Can't wait to see and hear hundreds of drones in the sky.

Yea this was one of my first thoughts too, drones are incredibly loud, even if its just one every few minutes it'll be unbearable. There are ideas on how to make them less loud, like putting the props inside the drone, but I suspect it will be very difficult to make them completely silent without flying really high.

Anybody know where this was filmed?

still gonna get shot down in texas

They are going to have to go back and update all the movies that attempt to guess the future.

Apparently there are going to be a bunch of drones flying around delivering to people in the suburbs.

Looks very cool but in the real world so many things can go wrong, I am sure people are going to intercept and either damage or steal the drones.

Haha. You must live somewhere pretty crazy if you expect that much violence and theft. People don't randomly attack everything. Sure, there'll be some losses, but that's true of any business, plus you'll have the theft on camera and just stop delivering to those neighborhoods. Compare the cost of drone loss to UPS' cost due to accidents. It would take an insane amount of effort to steal that many drones.

Yes I live in a low-incoming neighborhood and everything gets destroyed by unsupervised kids and teenagers that run around after school before parents get home (and some even after parents are home).

But I also live near college student housing and they pretty much do something similar with their drinking and partying.

Just like UPS cannot leave packages for me here, I couldn't get drone delivery either.

Not everyone can afford or wants to live in suburbs.

Low income areas also conveniently (for Amazon) have lower amounts of money to spend, so they are a much less desirable location to provide a service like this. They wouldn't have to provide this service everywhere for it to make sense to do it in some places.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact