Google's delivery drones combine a multirotor design with fixed-wings. It seems like an eminently sensible design to me; fixed-wing operation will be quieter and more power-efficient, allowing longer travel distances (unlike the early Amazon Prime Air videos [1]), while hover capability avoids the need to parachute the load (unlike the current Zipline drones [2]) and provides a hover-and-land-almost-anywhere capability I imagine is useful.
Presumably the winch-down-the-load design reduces the risk of a user putting their hands into the blades - not to mention reducing the risk from customers choosing an inappropriate landing location; I imagine the drone can also jettison the winch cord if it gets caught on anything. It precludes delivering directly to apartment windows - but AFAIK no-one else is trying that either.
It's a pity they haven't released clear info on these drones' range, round-trip time and weight carrying capacity. Which are of course critical to working out the profitability of a citywide coffee delivery service.
Mount Tennent in the centre background at 0:10 in the video. Tidbinbilla Deep Space tracking station is a bit off to the right (and behind a range of hills).
I'm frankly surprised that they've chosen Canberra (the nation's capital) as a launch city. The degree of nimbyism in Canberra would be very high with a lot of people who are very good at lobbying and political action (it's their job). In the case of this particular project I see that as a good thing. I think delivering coffees by a noisy drone that overflies urban areas is an incredibly stupid enterprise.
They probably chose Canberra as it has no state government, which is an entire monolith of bureaucracy to avoid, and in the NT they'd probably just shoot at them ;-)
Easy layout, yes. Fairly flat? No. Adelaide, Melbourne and Perth are far flatter.
Canberra is situated in a horst and graben [1] landscape with multiple prominent ridges separating the various urban 'valley' sections of the city.
Don't get me wrong, I can see some of the reasons why they may have picked Canberra. Generally a population with a high percentage of early tech adopters. Houses set back from the road with easy landing driveways at front. Power lines run along the back of property lines, not along the street. Etc, etc. Just I think they underestimated the resistance they are very likely to experience and also underestimated the danger of annoying the actual lawmakers who live there.
It's to keep the drone as far away from landing hazards as possible - you don't want someone accidentally losing a finger to a rotor or the drone crashing through someone's window or landing in a location it's likely to get stuck.
Yup it's a pretty good idea too. Pets, odd structures, children, terrain type, puddles, water run off areas.
Not to mention you need to account for the wind. Drones can actually fly really well in wind, and while obviously they won't be flying in strong winds, you never know if there'll be a gust.
In rural/suburban/residential districts the airspace is pretty darn clear above the treeline/buildings. Assuming you aren't next to a large airport, there is little to no air traffic, there's more wind but there's plenty of room, etc.
I couldn't tell from the video, but I'd assume they drop the entire string to prevent it from being snagged.
I saw the winch in the video and thought "what if someone grabbed the rope and used it to pull the drone to the ground", but then I saw the comment speculating that the drone could jettison the winch.
I predict that when drone delivery services are commonplace, there will be a community of drone hunting enthusiasts.
Delivery trucks are hard to transport and store once they're reported stolen. It'd be more sensible to just steal the packages on the truck.
The drone is "small", can be transported and stored without it being visible, so it get's lots of points there. Also I'd be concerned about stupid people/assholes and the shiny new drone.
That said, Cameras+GPS, not to mention you don't want to steal the drone if it's delivering to you because it'd be kinda obvious.
A better example could be electric scooters, which are indeed stolen sometimes, but apparently not enough to kill the various operator businesses (lime etc).
You know, I thought of a bunch of reasons why drones would be more likely to be stolen, but it never occurred to me that delivery trucks are accompanied by people.
> Is it possible to use a drone with a balloon on top to help it lift more?
That's essentially a blimp. Due to square/cube scaling, blimps have less lift and more drag relative to their size at small scales than large ones. The extra power needed to overcome the drag would by far negate the power savings from reduced lifting requirements.
Also, helium is for all practical purposes a nonrenewable resource. One drone wouldn't use a lot, but a few million refilling every day would put a dent in the world's helium supply.
I think having what essentially would be small robot controlled fire-bombs flying through the skies would be a tough sell. Even if you managed to make them as safe as possible it's not exactly the kind of thing people would be eager to approve.
It's worth noting that hydrogen is orders of magnitude more energy-dense, by mass, than lithium-ion batteries.
As a benchmark: the DJI Phantom weighs roughly 1kg, and its battery can store about 80kJ of electrical energy. I found a NASA study [1] which estimated that the energy released when a li-ion battery combusts is roughly 2x the usable energy capacity. So let's say that a battery fire would release 160kJ of energy.
In contrast, a balloon capable of lifting a 1kg drone would need about 830 liters of hydrogen, with a mass of 75 grams. That amount of hydrogen, if burned in room-temperature air, would release about 9MJ -- roughly 50x as much energy as a battery failure.
(Of course, most of the hydrogen would burn itself up quickly and harmlessly, assuming the balloon isn't flying near any flammable objects.)
Plus assuming the drone is above the tree line and not next to a tall structure, the actual "fireball" should move fairly skyward and not pose much of a fire threat? I'd be more concerned about the falling drone/battery.
It's not only direct contact with the fireball you have to worry about, a large amount of the energy in the explosion would be coupled into a pressure wave and infrared radiation.
What do you think gasoline powered drones are? Agreed that hydrogen is perceived as being much more dangerous than it is because of the Hindenburg disaster.
But (and I have no reason to think that this is true) if it turned out to be commercially viable, I'm sure that perception could be changed.
Agreed. But there's a perception issue that arises from flammability that requires political skills I don't have. So hydrogen feels like a difficult solution to me. Example headline: "Flying Bombs, Oh My!"
If there were a mini-compressor-pump aboard the drone, is it possible that it could inflate/evacuate a balloon on demand during hover mode? Is the economics of such a thing
even possible?
Assuming there weren't already problems with the balloon idea, what you'd have then is a pressurized vessel, the pump, and of all the compressed helium weighing you down for 99% of your trip just to save a tiny bit of energy (and I'm assuming a day with zero wind here) during the offloading procedure which probably doesn't even last long enough for the balloon to fully inflate.
Also, as far as I know, the cheapest/lightest material to make the balloon out of would be mylar, which is conductive and not something you want around powerlines.
You'll experience the same scaling issues with whatever it is you expect to inflate with. You're going to have a lot of volume dedicated to the pressure vessel and not so much in the pressure vessel, because of scaling issues.
And I'm discarding the pump to put it back, assuming we're just going to vent to the atmosphere when we're done. Things get worse if you want to pump it back in.
No that doesn't really work. High pressure helium pumps are expensive, heavy, and require a lot of maintenance. Plus they require a lot of power to operate. And they produce a lot of waste heat.
Not saying this idea is worth the complexity (sounds over-engineered), but I am pretty sure you could make like a 1"x2" can of compressed helium attached to a 3' balloon with a simple valve to pop open and inflate the balloon. Think of those little air cans that you use for refilling bike tires on the go. Only weigh a few grams, holds a lot of air.
Hadn't heard of Zipline. And when I read it I imagined a simple cableway with small autonomous cars zipping along the line, possibly with branching points. I suspect it would be way cheaper to set up as an infrastructure and much more reliable than drones.
We're talking about drones which can serve an area of 7800 sq miles (20000 km²), going at over 60mph (100km/h), across rivers and mountains inaccessible by vehicles.
I don't think there's any way you could even install enough poles and cables to cover such an extension, let alone keeping them functioning without permanent maintenance teams trekking across the peaks. Plus all that would fail exactly when you need it the most: after a big storm, when the roads are impassable and even walking is difficult.
Yes, I was thinking more of an urban context, there are surely places and situations where a real flying drone (basically a model airplane in this case) works better.
I was merely trying to suggest that maybe we are too dazzled by the technology (flying! straight to your garden!) to see simpler and more effective solutions.
Drones don't really require infrastructure though? They don't require any setup at the physical locations, and they have a pretty obstacle free path above the trees.
Zipline would require infrastructure setup on the scale of roads/telephone lines to be useful wouldn't it?
Yep. But it seems it would be rather cheap to set up an infrastructure made principally of steel towers and (thin) cables, and there shouldn't be many technological challenges. It would be probably faster, more reliable and energy efficient than drones, allowing for heavier packages. Of course at the cost of having to build an infrastructure, and maybe of a smaller, fixed number of delivery points.
> But it seems it would be rather cheap to set up an infrastructure made principally of steel towers and (thin) cables, and there shouldn't be many technological challenges.
Just huge logistical challenges with right-of-way, construction labour, tearing up and making good the {sidewalks,lawns,gardens,driveways,etc}, actual acquisition of poles, ongoing maintenance and trimming duties, etc.
Greenfield installation of infrastructure in public areas is incredibly complex.
Yes, and I imagine it would be really difficult, for good reasons (mostly esthetical), to convince a city to install poles and cables for package delivery. At the same time, we've seen drone trials going on for some time, and the idea of having expensive, autonomous flying machines, on a very limited battery life, carrying weights over our heads and delivering them possibly without even landing but just hovering and lowering a winch, well it seems pretty implausible.
A Google wing drone probably costs a few thousands, can deliver two cups of coffee on range of about 1.5km, in extremely controlled conditions, for 100 suburban homes. And has already pissed off people in the neighborhood. Imagine how it would work inside a city center, and how it would scale. And how would you adapt it to homes and buildings that have no garden and no roof terrace?
I actually have a drone I fly around freestyle for fun, after seeing how truly versatile they are, not to mention how cheap, I'm really not sold on constructing more towers.
The EMax Hawk 5 is 250$ and while I don't know how much it can lift, it's actually a VERY decent amount, if we're just talking about a single person's mcdonalds then no problem. The expenses come from making it more robust, making sure it has signal, etc. But those issues are mostly already solved, especially with a winch the drone can stay above the tree line, a single radio tower would have incredible range - or they might be able to get away with existing 4G towers.
Now you're just down to battery life, and noise. The hybrid fixed wing/multirotor system seems to at least partially address both.
Battery life. And noise. And carrying anything substantial. And adding all the safety mechanisms that prevent it from killing someone in case of malfunction. And mechanisms to prevent it from being stolen or damaged. And autonomous navigation in case of unforseen obstacles (a construction crane?). And the ability to either land or negotiate a delivery point (imagine an urban road, with tall builings, walkways and traffic). And the ability to coordinate with other autonomous drones (avoid collisions, etc.).
I do fly a drone for fun as well. Last Christmas I had the idea of flying it over the frozen surface of a mountain lake, not far from my hometown. Well, about 30 seconds in, a pretty angry buzzard made a straight line for it and crashed it. It weighs about 2.5kg, it crashed on the ice from about 10m. No major damage, but retrieving it from there was, well, tricky :)
> Wing said the feedback obtained during its trials had been "valuable" and it hoped to "continue the dialogue".
That's code for they have asked their lawyer to see what was legal and if they can simply ignore the complaints.
We continue to wage war on silence. The other day I was far the heck out there on a trail and the only things I could hear were squirrels birds and a few other things. My soul was at peace. We can't have that in cities or even suburbs any more. Instead we get BUMP BUMP BUMP from passing cars. The level of anxiety we get from cities is insane and people don't really even understand this until they are 60 years old. I'm 38 and I understand it.
I don't know where you live. But if you were in the US then pretty much all of the land is unpopulated and quiet (except maybe for agriculture purposes but still).
What you are looking for is city infrastructure with green scenery. That's possible but the price tag is going to be astronomical. You'd have to provide everything yourself (Water, Electricity, Gas, Food, etc...) Try that and you'll find out what you prefer living close to a city.
> What you are looking for is city infrastructure with green scenery. That's possible but the price tag is going to be astronomical.
In a car-centric culture like what dominates in North America, maybe. I've been to a city or two, though, that accomplished the hat trick of high population density, low noise levels, and nice scenery. Brussels, for example.
I'm pretty sure the magic trick is to convince your citizens that public transportation doesn't have to just be a last resort for poor people.
Indeed, I prefer noise. Silence makes me feel ill at ease, like nobody's around, like something terrible could happen and I have no recourse to anybody. I don't feel safe in silence, but whenever I tell anybody this, I'm poo-pooed as being silly.
If only there were quieter places, out in the countryside, where you could have Really Uninterrupted Relaxing And Leisure - hey, we could even call them that: R.U.R.A.L.
An available substitute is a great reason to allow a particular service to not be provided in a particular location.
If rural environments weren’t accessible it becomes much more important for cities to be more quiet. Because rural environments are available, cities can trade some silence for some pretty extraordinary economic advantages.
Building skyscrapers and moving millions of people around every day is noisy work, but it provides one heck of a boost to the GDP.
If cities can be quieter for free, they should be. If being quieter costs cities one of their inherent advantages than people who don’t like that can choose a different environ.
Sure. But we still should consider both sides of the equation. Even in a city, not every loss of silence is worth every gain of productivity or other city advantages.
Also, in this case we are not exactly talking about Manhattan, but about houses with gardens in Canberra.
It's difficult to imagine that this is ok from a privacy standpoint. I mean yes, the claim is that there is a "low-resolution fitted for debugging purposes", but why would Google stop there?
There are now cameras flying around Canberra and there is seemingly not a single way for a civilian to differentiate between a Google delivery done or a fake drone used for spying. There will be no way for people to opt-out of this.
I see this will end up similar to the way people who send emails to Gmail users cannot opt out of Google being able to use their personal information without consent. You will be on camera whether you like it or not.
Canberra also supports large native bird populations, this can't be too healthy for their stress levels as this scales up.
It is surprising to me that this is the straw that broke the camel's back. There are billions of Android users that appear to be browsing the web on their phones but are actually taking pictures of you (you personally!) and uploading them to Google. There are planes flying around taking pictures. There are drones flying around taking pictures. There are satellites flying around taking pictures.
The cat is out of the bag. There are pictures of you. Not sure why that should stop drone delivery (or self-driving cars for that matter, which guess what, have cameras on them).
There are not satellites taking photos of people as frequently or as in close proximity as these things will. Google would literally be able to identify even how healthy you’re looking by image classification.
To you’re point about Androids, yeah, it’s not healthy or good that the thing is effectively a data mining apparatus. So now, even if you decide to leave the phone at home to get some privacy, now you’re going to have to bring a disguise too.
I’d agree that it would have gotten old if this was an isolated past of Google but fact is that Google shuts downs products left and right even today, recently Google Inbox.
People will keep making fun of Google for this as long as Google keep throwing products at the wall that they don’t intend to keep supporting.
Googles inclination to deprecate products is common knowledge at this point. What value is added by restating the obvious in every Google related thread?
Except, you know, this is Wing, a separate entity.
Besides. It almost sounds like many people would like them to stop releasing products all together. Market research is not yet at a point where you can predict if something will be 100% successful.
I think the problem here is with Google's definition of success - they're cancelling products with many millions of satisfied users.
I'm not saying I have a perfect solution that would satisfy everyone. But, in the case of Reader or Inbox, for example, I think that they could keep the business going with about a couple of cents from every user. I feel that a company with as much power as Google should really do more to stay on its users' good sides.
I think it's the beginning of a relevant discussion thread on any Google service. At this point it's such a significant problem with anything Google runs that we should be talking about it every time they launch something.
I personally expect that the first successful drone delivery service will be implemented by some sophisticated drug dealers because unlike with all these drone delivery services, they would need to be quiet and reliably deliver products instead of publicity.
“There’s not a prison not fighting this issue in the country,” says Sean Ferguson, New Technology Project Manager at Georgia State Department of Corrections. “If they say they’re not facing it, they’re just not admitting to it. So it is an issue.”
I know it's a small drone, but I own lots of small drones and none of them are that quiet, and the tech will likely be applied to larger drones over time.
Don't drones have to be controlled from a certain distance? Most prisons around here are in remote areas, so if you see car pulled over on the side of the road near a prison, maybe they are flying a drone to deliver contraband.
That's a pretty slippery slope, don't you think, if law enforcement started pulling up on all stopped vehicles within X distance of a correctional institution?
It reminds me of the issue where constitutional protections can be selectively removed within 100 miles of the US border by border agents.
Legally they're not supposed to fly past a certain distance from an operator, but if you don't care about the law you could easily operate via cell towers, or even autonomously via a predefined GPS coordinate based route.
One needs to restock these dead drops, right? Maybe a distributor can bypass all the small time street dealers and create an empire whose only distribution channel is drones.
You must be from US - it seems to be lagging behind in that area. Here, there are not street dealers anymore, deaddrops are how end users buy it. And hiring a stream of low-skill couriers that end up in jail after 2-3 months of working for anonymous boss is way cheaper and safer than flying drones from your own location.
At least around here (in US), while street deals are still the common among most people, USPS is likely the one of the biggest/safest methods now. They require a warrant to search domestic packages, not to mention the most common "illicit substance" is now commercially legal in many states...
No one talks about the noise. Every time I read about drone delivery, this major aspect is omitted from the discussion but I believe it would be a nuisance to have drones droning around all the time making a huge amount of noise.
Airplanes make a lot of noise and it is still a hot topic in neighborhoods adjacent to airports.
Little hobby drones make some noise but when you’ve got a drone lifting 5kg of weight, it’s going to make a tremendous amount of noise.
Noise is already solved in the one they're using in the commercial service, per the article. Politics also were cleared, as the aviation authority there gave them permission. So yes, financial feasibility remains.
Fair, but it's at least solved to a degree with they have permission to use it commercially. There could be other standards for what's considered solving the problem, but that seems like a reasonable one.
I've though that the original Project Wing design (flying wing that transitions 90 degrees for vertical hover) would have been more efficient. Or a larger version of the E-flite X-Vert VTOL:
I suspect they're still headed that way long term, and just went with this design to get testing ASAP because the pusher configuration has more challenges.
The flip maneuver is difficult to integrate with their sky crane style system. You absolutely don't want to tilt your payload because that ruins lots of foods (coffee, pizza) so you either need some sort of pivot mechanism for your payload attachment smack dab in the middle of the craft, or you have to do some daring aerobatics to finish retracting the crane as you begin low speed level flight.
Their noise problem is almost entirely because of those tiny propellers - A pusher wing with a pair of much larger (10+"), slower props could be pretty quiet.
You are semi-correct. The tail sitter design was abandoned because it was too difficult/unstable to control while using the winch.
The core team that started Wing quit X awhile back and formed a new company, Skydio, which has built an incredibly cool drone for the purpose of high-tech narcissism:
Interesting, the FAA has announced the first drone delivery air carrier approval in the US will happen in April. I would have thought Flirtey would win this race but according to the NYT the only application listed is from Google Wing.
> Rakuten has been offering drone delivery in Japan since 2016 and unmanned vehicle trials since 2018. It said that working with JD — which claims to have racked up 400,000 minutes of delivery flight time — will “accelerate the development and commercialization” of its human-free last-mile delivery efforts.
"They will not be allowed to fly over crowds or main roads."
I know this is specific to the current set of guidelines, but if I imagine this service launching in my part of the world I'm unsure how it would even work with this requirement. To go from one side of my city to the other you have to traverse a number of main roads. If you're setting up base stations all over the place to launch from to keep from breaking across main roads I can't see how this solves the distribution problem (though it may solve the "we have to pay humans to deliver packages and takeout food" problem...)
What's the use case besides delivering to remote areas ?
I can get Zipline's mission [0], but personal deliveries ... ? A UPS truck with hundreds of packages surely is more efficient than an army of drones in urban areas.
> "Wing is developing a new method of transporting goods that’s faster, cheaper, and more environmentally friendly"
What's the advantage compared to an electric van ?
> 27% of greenhouse gas emissions come from transportation.
I mean, yeah sure, but changing the last 3km of the delivery from van to drone won't change the fact that the coffee beans or new iphone you just bought came from the other side of the world via shipping containers which is 99%+ of the pollution associated to the delivery. [1]
Your medium link doesn't support the 99% claim. There's no way the last mile is less than 1%. Strikes me as obvious that cargo ships are vastly more efficient per pound than a FedEx truck.
I imagine this would be used for Immediacy solutions. Amazon Prime Now being a great example of which.
It would cost more, but per order it's probably cheaper than a guy on a bike given guy on said bike + insurance + overheads will take a bigger cut than the drone will.
More often than not takeout delivery services will drive with gas-fueled cars, and reducing emissions is one of the big points the article and video [0] touch on.
One advantage could be speed - drones will allow you to move faster for smaller orders. Just like you have the shipping container -> warehouse -> trucks -> local warehouse -> smaller trucks -> home, you can extend this to respond faster to customer requests, as well as not involve any humans in the whole last mile delivery system (avoiding costs etc)
3-4 tall pylons with heavy duty winches and kevlar cable linked in the middle to hold a winch system that can lower a delivery box anywhere in the city.
1/4 inch kevlar rope weighs 2lb per 100ft, therefore a 2 mile kevlar rope would weigh 210lbs, but has a load bearing capacity of 9500lbs. So this system would allow accurate delivery anywhere in a 2mile x 2 mile area.
Potentially these could be coupled to allow transfer between areas to allow deliveries to be passed along a chain.
Multiple winches could be stacked on pylons to allow multiple sets of ropes, with winches which are near to delivery/pickup points moved downwards and to lower rope tension to avoid other winches/ropes bumping into them.
the infrastructure cost is way too high. Not just the cost of materials and labor, either. I'm talking social, political, and environmental costs. people would fight this left and right.
Power lines are already considered an eye sore in many places and are required to be underground. good luck convincing people that a power line with a bunch of moving boxes on it is going to be any better. its essentially a gondola for packages and no one wants that in their backyard.
You'll have large shadows defined by where buildings are in relation to the pylons and the shuttle would not be able to descend below the roof line in many areas outside of a subdivision with low 2 story houses depending on how tall the towers are. In addition you can't really do this anywhere there is going to be air traffic with the current systems as you'll need to shut it down whenever a helicopter wants to fly through.
I'm wondering what exactly Google will use this home delivery by drone tech for? Their Google Express service is nice but it's not their main business and this makes a lot more sense if it was launched by e-commerce giant like Amazon. I'm just wondering how Google will convert this tech in a viable business?
The main area Assistant is worse at compared to Alexa is shopping. They kind of need to have their own e-commerce platform. You could (eventually) combine that with drone delivery.
When in college, I once delivered to a 3rd story balcony via ladder. Anything is possible if you hire the right people. Unfortunately, this service requires a large spot in your yard. Sorry apartment dwellers.
People with the yard space will become hubs (for a relatively small fee) for apartment dwellers. Or, the drones will deliver to the roofs of the apartment blocks instead.
I can envisage a future where you have huge drones with landing pads that fly between hubs in a set amount of times.
This will then allow individuals to request a drone from a 3rd party. Load up the package and program the drone for the final destination. The drone flies to the LZ drone where it lands and waits until its main destination and then resumes flight until it delivers the package.
The drone could have a wallet where it takes the initial payment from the customer. Which a % goes to the drone operator, when the drone lands a % goes to the LZ operator and any other stops.
Having essentially P2P delivery system will be a total game changer!
I think with today's technology there's no need to build a card reader into the drone itself. It can just be handled online prior to delivery (or after successful delivery).
Google's delivery drones combine a multirotor design with fixed-wings. It seems like an eminently sensible design to me; fixed-wing operation will be quieter and more power-efficient, allowing longer travel distances (unlike the early Amazon Prime Air videos [1]), while hover capability avoids the need to parachute the load (unlike the current Zipline drones [2]) and provides a hover-and-land-almost-anywhere capability I imagine is useful.
Presumably the winch-down-the-load design reduces the risk of a user putting their hands into the blades - not to mention reducing the risk from customers choosing an inappropriate landing location; I imagine the drone can also jettison the winch cord if it gets caught on anything. It precludes delivering directly to apartment windows - but AFAIK no-one else is trying that either.
It's a pity they haven't released clear info on these drones' range, round-trip time and weight carrying capacity. Which are of course critical to working out the profitability of a citywide coffee delivery service.
[1] https://www.youtube.com/watch?v=MR9PoBAssw0 [2] https://www.youtube.com/watch?v=bnoUBfLxZz0