Hacker News new | ask | show | jobs
by nomel 1776 days ago
My naive assumption is that regulations (appropriate or not) is what prevented this from getting off the ground.
5 comments

Yup. Zipline in Rwanda and Ghana have made over 150,000 commercial drone deliveries. And have done so safely. (Their autonomous winged drones have a range of over 300km on a charge.) https://www.morningbrew.com/emerging-tech/stories/2021/06/30...

Only token demonstration projects in the US, unfortunately.

They've found a niche there where drone deliveries can work and be worthwhile. Getting lightweight but vitally important payloads (medical supplies) quickly to locations hard to access by road.

Amazon is in the business of delivering bulky and heavy packages, often of low value and importance, and poorly-but-cheaply packed into oversized cardboard boxes. I can't imagine a very large percentage of Amazon deliveries would be in any way suited to drone delivery, especially when you've got to limit it to destinations with a suitable landing/drop zone, within a certain range of a depot but away from dense urban areas (too high-risk to operate in)

Yeah, I live in a relatively rural part of the US and can drive hundreds of pounds of stuff at 60 miles an hour to almost anywhere within a few hundred miles.

That doesn't eliminate rapid delivery of small packages as a business model, but it really constrains what is worth shipping using a drone that can carry a couple of pounds.

Right. But a lot of Amazon deliveries are single packages, and probably most of them are within the ~4 pound weight limit of current Zipline drones. 80mph, straight-line one-way delivery without requiring a dedicated human driver is pretty effective for rural delivery (and the drones have a range of 300km per charger, currently). But it really is more like Starlink than fiber. It's better suited to underserved rural areas where it's impractical to do single-package delivery.

And I think the current ~$15-20/delivery price for Zipline is too high to be terribly practical for most items, but in a rush (and given 24/7 availability), it's absolutely worth it. I can think of many personal instances where a $15-20 delivery charge would easily be worth it to trim hours off of delivery time.

I think at this point it's likely safer (i.e. to pedestrians, etc) than delivering the same item via truck. I really do think this has a future, and I'm glad that they've reached a sustainable level of operations in Rwanda and Ghana.

My former startup got Jersey City to draft legislation that fully paved the way for delivery drone to operate in the city, they even went as far as to have the FAA clear their draft legislation, basically on the premise that drone companies would take it seriously, talked to Prime Air and they just shrugs and said sorry, not interested.
sure

> UK regulators also fast-tracked approvals for drone testing, which made the country an ideal testbed for drone flights and paved the way for Amazon to gain regulatory approval elsewhere.

It is literally regulations that keep unproven aircraft grounded, yes
Only if they’re uncrewed. Any idiot can legally fly an ultralight experimental aircraft with basically zero oversight. I don’t think you even need a pilot’s license. But God forbid you try to do the same thing without putting a pilot at risk; that’s effectively illegal.
That makes some amount of sense. If you put the ultralight pilot in the craft, they literally have skin in the game. Further, ultralights cannot be flown over “congested areas” of a town/city.

Trying to apply those rules to drone deliveries while ignoring those two pretty fundamental elements doesn’t strike me as evidence that the FAA is being arbitrary or capricious.

Which actually makes sense.

When you're risking your own life, you'll likely be at least somewhat careful.

When you're just flying an empty test vehicle you'll be much more reckless, and thus way more likely to crash it on other people's property or other people.

Hence, much more tightly regulated.

Part 103 aircraft don’t need a license of any sort. There are of course a bunch of regulations about how/when/where you fly it, but you don’t need any official training or certification.
> Any idiot can legally fly an ultralight experimental aircraft with basically zero oversight. I don’t think you even need a pilot’s license.

But they have limits:

No passengers allowed.

No flying over towns or settlements.

No flying at night or above (or in) the clouds.

No flying in airspace around airports with control towers and certain other airspace without prior permission.

No commercial operations (for hire) except instruction.

Ultralights must yield right-of-way to ALL other aircraft.

Idiots flying ultralights are a problem that’s going to self-eliminate quickly.
The only hurdle was that the drones had to be so heavy to handle packages that it put them in a higher weight class subject to more rules. But other than that, the article mentions that the UK and the US regulatory bodies have both approved the project.
I wonder what the stats were on weight of packages vs required drone weight. Ie where is the cutoff of package weight, to ensure that the money they invest in drones would affect X percent of sales?

Eg i imagine if they made drones for a package weight of, say, 10 ounces, they could.. hopefully, make smaller drones and perhaps have an easier time with regulations. However the package size limit is so small that the drones aren't likely worth it.

So as we increase package weight, we include more packages and percentage of sales... but also increase drone size and regulatory issues.

I imagine there isn't a "sweet spot" otherwise they would have done it. Nevertheless i'm curious what that axis looked like.

I don't know about the UK, but in the US the FAA required that each drone be operated by a licensed pilot, which changed the economics.

There's some talk of autonomous software being granted a pilot's license, so that might shift things again.