Mercedes-Benz conducting the biggest test using drones to ship everyday items | Hacker News Reader

Mercedes-Benz conducting the biggest test using drones to ship everyday items

bloomberg.com

53 pointsflinner44 comments

I can imagine drone delivery's an effective solution to the last mile problem for automated delivery? Have a self-driving truck navigate through a neighborhood, the drone takes off with the package and delivers it to the doorstep and returns to the truck. The truck can continue driving around as drones rejoin it down the road.

Now we have both pieces of the equation: how to get to the house without paying a driver, and how to get to the porch without paying someone to walk and carry.

How many jobs would that “repurpose” in delivery if you add a automated truck? Energy and materials to build and maintain drone fleets, noise, all seem like issues, since this is added on to the existing truck.

You make a good point. Would those steps actually lead to cost savings? Driving a truck and delivering packages requires less specialized skills than maintaining the fleet. How do delivery companies handle maintenance on their human-driven trucks? It is contracted out? If the company can run three trucks per staff member vs. one per truck, wouldn't the savings be worth it? Are they going to pay this new staff member 3 times as much as a driver?

What is this going to sound like when cities have hundreds of whining drones flying around? Are we just going to have to get used to the noise pollution from the drones so people can get their coffee delivered faster?

What is this going to sound like when cities have hundreds of whining cars driving around? Are we just going to have to get used to the noise pollution from the cars so people can get their coffee delivered faster?

-------

This actually might lower the delivery times and delivery cars on the road which is a good thing! Less traffic, less air pollution, faster delivery times, less jobs (not a good thing tho).

It's true that we already deal with annoying vehicular noise, but that's hardly an argument in favor of repeating the same mistakes.

I'm not convinced it would be anything like a net gain. I'm not a fan of car noise, but a delivery truck carrying hundreds of packages makes less noise than a drone too small to even carry one of them.

Hopefully it's possible to make these things much quieter....

No, they won't be much quieter. The noise is a function of the shape of the propellors, their diameter and the speed of the air flowing through them. The typical drone blade has to run at 7K rpm in order to just hover the drone, 12K and up to maneuver. That's very much in the audible range and the speed with which the air flows through and the fact that the air flows downwards through them pretty much guarantees a very loud presence.

A couple of these at all times and you'll be ready to start target practice.

We should design them to have owl like wings. Owls fly very quietly.

You should definitely watch this:

https://www.youtube.com/watch?v=lVo3wEkyci4

and

https://www.youtube.com/watch?v=nnR8fDW3Ilo

Also, while birds are really good at using their wings to glide and provide propulsion flapping designs that need to lift much more than their own weight have so far not worked very well. So I don't really see them as viable for packages.

Interesting. Seems like some birds can lift a lot of weight though like a hawk carrying a rabbit.

Yes, but for a given weight in flight so far nature seems to be doing much better than batteries or fuel when it comes to carrying capacity. That's not to say that will not change, but for now biology has the edge, especially when you compound the problem such as limit the noise emitted.

True, but imagine if these drones handled food/grocery/retail delivery, the amount of cars on the road would drop immensely. As long as the tradeoff is 1 car -> 1 drone.

Trading off one car for one drone would result in a massive increase in noise. To be a win, you'd need each drone to replace hundreds or thousands of cars, and right now the equation is the other way around.

Cars are getting quieter all the time, too. Tire noise already dominates for most vehicles at city speeds, and EVs improve the picture even further. Diesel trucks are often pretty noise, but they're mild compared to a drone, and electric trucks will drastically cut down on noise.

Not to mention in most places, highways (the places where cars make the most noise) are far away or acoustically shielded from where there are people.

You’re assuming nothing else would fill the gap left. It’s not a static system.

So progress is hell on earth and technocrats are cheering it up?

It would replace a single vehicle with dozens, if not hundreds of drones, all traveling longer distances than that single vehicle.

If we take the example of a delivery truck full of many packages, it seems to me that each drone would travel a shorter distance to their single destination than the truck that had many destinations.

When a delivery truck stops in front of my apartment building, it delivers a dozen packages.

A drone would have to make a dozen trips to do this.

I have to say, the noise generated by the UPS truck stopping-and-going on my street is... Insignificant compared to even a single consumer drone buzzing overhead.

> What is this going to sound like when cities have hundreds of whining cars driving around?

Car noise is a big issue.

We had to start designing cities around cars.

Automatic pilot for a flying drone from A to B is much easier than one that moves on streets or sidewalks, but I hope we don't let these flying drones cover our cities with noise. If automatic cars work out, maybe a system of quiet, electric trucks that stop at your house/apartment and then a flying drone from truck to the front/back porch (or open widow? That would be cool for condos/apts) would not add much more noise to the urban environment.

Additionally, how will the drones be secured? A baseball could knock these things out of the air. It's not like with a delivery person where there's a lot of risk involved trying to take a package.

A drone large enough to ship a package will probably not be taken down by a baseball. It's also going to be difficult to throw a baseball at a moving drone. You'd also have to know when the drone is going to come by, because most of the time, they would be flying high, except when near landing. Finally, security cameras.

A drone large enough to ship a package will kill you when it falls down for whatever reason.

I don't see this whole concept working simply due to (re)liability.

well the concern I am curious about how they address it is, how will they route these drones? How high up do they need to be when crossing private property before they are not trespassing. If any have cameras attached how do we verify they only record where appropriate? if routing down existing roadways how do they fare in areas with large numbers of over head wires?

my personal concerns from above are trespass and privacy. besides someone should easily be able to come up with a means to track them and nab some packages, with drones if there is enough money in it.

Most drones would likely be carrying low-value items. The risk would probably not be worth it. Each time you try to nab a drone, you are exposing yourself to the risk of being arrested and prosecuted. If security is a huge concern, keep shipping high-value items by people. Program the drones to fly in unpredictable patterns, etc.

> The risk would probably not be worth it.

There are the parts the drone is made out of. A drone capable of lifting a package would be worth a pretty penny in parts.

I had one fly low near my top-floor unit recently and it was way louder than I ever expected it to be.

100 years ago, you would have said "What is this going to sound like when cities have hundreds of whining cars driving around? Are we just going to have to get used to the noise pollution from the cars so people can get their crackerboxes delivered faster?"

And they would have been right, which is why cities are now (finally) trying to curb that plague.

I suggest a free-market approach: let these companies purchase the noise rights of each affected inhabitant. Then Coase can decide if having a crackerboxer two hours sooner is worth the cost.

Cities aren't cracking down on cars because of noise, they are cracking down because they are a very poor use of limited space.

I'm curious if you think people who are,say, proselytizing on the sidewalk should need to buy noise rights from people who can hear them as well?

We seem pretty used to the noise of trucks, which are the loudest thing in my neighborhood.

Who had the bright idea to put a drone video up top that has nothing to do with Mercedes? The company in the video isn't even mentioned in the article.

Stay weird, Bloomberg.

That's the reverse of Mercedes' previous robot delivery scheme. They did a test with Starship Technologies delivery robots, where a van was filled up with the robots, it went somewhere, and the robots fanned out to deliver things. That would make sense for dense areas like apartment buildings and offices, once you get doors and elevators to cooperate.

This new test delivers stuff by drone from a distribution center to the vans, which then make local deliveries with humans. That seems backwards. The drones are doing the heavy lifting, something trucks do very well.

What's the failure mode for quadcopters, can these land safely with loss of one motor? Definition of "safely" would ideally consider human casualties.

Well, this happened a couple of years ago in a World Cup skiing event.

https://youtu.be/xeviAWB0i4Y

That's an incredible link, thanks for that.

That someone would be skiing, whatever, 100 miles per hour and might then die because of a falling electronic lawnmower really highlights seen vs unseen risks.

Shades of The Truman Show. That was bloody close.

Biggest seems to be a superlative with no basis. Perhaps its biggest test?

All this drone delivery stuff is theatre until they start building obstacle detection and avoidance into them. Yes, drones can take off from point A, then fly up, over to point B, and land. We already knew this.

Even modern consumer drones have obstacle detection and avoidance built in

You can buy $300 drones with obstacle detection / avoidance. Hell, I even recall a service on reddit where someone let you fly his drone around New Zealand and it was completely unmonitored because the drone would not let you crash or hit anything.

Last mile is best served on foot, not by vehicle.

People wonder why the bees are gone or why it's gonna be a total wasteland for life in no time.

Reply on news.ycombinator.com