Amazon's Satellite Project Will Cost Billions, Jeff Bezos Says | Hacker News Reader

Amazon's Satellite Project Will Cost Billions, Jeff Bezos Says

bloomberg.com

57 pointsjason_zig60 comments

I assume there are a finite number of these swarms that we can deploy in LEO, both in economic terms as well as dedicating space and spectrum?

I’ve heard of three companies now that want to achieve this. I’m not sure if there’s anything particularly distinctive about one versus the other, other than SpaceX is the only one with the proven launch capacity to bring that much mass to orbit.

I have a lot less faith in Bezos pulling this off, but mostly it seems a waste to start an ISP just to earn Amazon shareholders a larger dividend.

Dedicating space? We have three deminsions to work with, but even if it were two, space around earth is bigger than surface of earth. The earth is very big and these constellations are small. The satellites are small too. At this low an orbit, not worried about long lasting junk due to atmos drag iiuc

Amazon shareholders do not earn a dividend. SpaceX is not public, but rather would enrich a much smaller set of investors (and particularly Musk). What’s your point?

Aren't SpaceX creating Starlink to generate funds to support other projects; in particular, their plans to create the rockets and infrastructure to ultimately colonise Mars?

i.e. not (primarily) to "enrich a much smaller set of investors (and particularly Musk)"?

Yeah, that was silly and disingenuous. We're talking about a guy who could have retired in lavish luxury 15+ years ago if 'riches' were all he wanted. He is Mars or bust and looks like he's put together a company with the chops to pull it off.

Fidelity has a biyearly traded SpaceX fund.

Isn't it just a small slice of many things held in a mutual fund? or there is actually pure spaceX fund/vehicle?

My point was that when companies look for literal moonshots to keep finding decent uses for their money, it’s usually not long after that the dividends start be paid out.

That is to say, by the time their Starlink-esque network would be generating revenue, odds are good AMZN is paying a dividend.

Yes, there is a finite limit, but the number is much higher than you would suppose.

First, in terms of spectrum, realistically all the satellites in a constellation can (and should) share a small number of relatively narrow frequencies. You can almost imagine it ma be the equivalent of one more cell tower in your area per constellation (at worse).

In terms of space, astronomy aside, the actual space these satellites are flying in are “huge”. As a perspective if you hold your thumb up to the sky, the distance “covered up” by your thumb (if I did my math right) is approximately 480 miles [1]

Yes things travel very quickly at those orbits, but the point is that even if we throw a ton of satellites up, they will remain fairly sparse.

That said, possible reflection and how noticeable they will be, that is a whole other question.

[1] assuming orbits of 366 miles up, higher orbits would be even farther apart

> As a perspective if you hold your thumb up to the sky, the distance “covered up” by your thumb (if I did my math right) is approximately 480 miles

What distance? The distance immediately behind my thumb is about an inch. The distance way back at the edge of the universe is millions of light years

The parent comment says in a footnote:

> [1] assuming orbits of 366 miles up, higher orbits would be even farther apart

I’m confused on why it would be a waste to make shareholders better off, regardless of the fact that Amazon does not pay a dividend. Isn’t this the point of investing? It would be great to make shareholders better off and even better if we could improve internet access while we are at it.

> mostly it seems a waste to start an ISP

I sort of see an analogy with airplane or helicopter pilots.

One of the most important things pilots need when they start out is flying time. And flying time costs money. They can be easily talked into things to defray their costs.

I suspect that launching an ISP is defraying costs of developing and launching rockets. $100 LEO hamburgers.

Gotta give it to them, Amazon is the only big tech company with a vision - the rest are just sitting on hundreds of billions of dollars in cash and no good ideas against which to deploy all those funds.

Google invested $900 million in SpaceX back in 2015 [1]. Not sure if that counts as "vision", but it seems like it was a smart way to get into this market early.

[1] https://spacenews.com/google-spacex-investment-is-900-millio...

Google attempted something similar with loon[0] and FB with internet.org[1] (IIRC, FB also had a few other initiatives)

[0] https://loon.com/

[1] https://www.wired.com/story/what-happened-to-facebooks-grand...

And with regards to Amazon being the only big tech company with a vision -- I hear Apple just redesigned the monitor stand. Visionary.

It's detailed more in your second link, but Facebook was planning to launch their first batch back in 2016, which ended up blowing up due to the failed SpaceX launch. So if it wasn't for that mishap, they may have had something operational by now.

As pointed out by others, Google has had visionary projects for over a decade. Just take a look at their X projects [0]. Here's a few:

- Verily is looking into improving medical tech

- Malta is looking at more efficient energy storage

- Makani is looking at innovative wind energy gathering

- Loon is looking at connectivity for rural/remote communities

- Foghorn is looking at creating clean energy from sea water

And of course, let's not forget about Waymo which started 10 years ago, way ahead of its time. Google also invests heavily in many industries. They fund multiple fusion energy projects [1][2][3], they had a million dollar prize to encourage people to get to the moon [4], and as pointed out above they are big investors in SpaceX too.

Meanwhile, other than Bezos' Blue Origin, I think this is basically a first for Amazon, and honestly even this feels much more of a business move than a "visionary project", considering there's literally 3-4 other companies trying to do the exact same thing (SpaceX, OneWeb, Facebook, Telesat).

[0] https://x.company/projects/

[1] https://www.popularmechanics.com/science/green-tech/news/a27...

[2] https://www.theverge.com/2019/1/29/18201220/nuclear-fusion-e...

[3] https://www.forbes.com/sites/christopherhelman/2019/05/21/th...

[4] https://en.wikipedia.org/wiki/Google_Lunar_X_Prize

Amazon also had press this week about drone delivery.

Quite a presumption that announcing specific developments today is a singular sign of a big tech company’s vision.

In a carefully oriented company, I think drips of vision are hidden in every executive interview and minor product release.

You only get major, obvious vision-level clarity once in a very long while.

This is a business that Apple should get into as well. They have the capital and Phones+internet would have great vertical synergy.

What they don't have is the rockets. SpaceX has cheap reusable Falcon 9s. Blue Origin is aiming to have a cheap reusable New Glenn in 2021. Apple would have to go with much more expensive traditional launchers.

Why would Apple not be able to buy Falcon 9 launches? SpaceX is in the business of selling rocket launches.

They could but I'm guessing SpaceX margins on rocket launches are ~50% (to make it economical in the face of risk of blowing up billion-dollar payloads).

Which means that if Apple pays $60 million for a single launch, it costs SpaceX $30 million to do it.

Or to put it differently: if Apple was competing with SpaceX, then every Apple launch would sponsor one SpaceX launch.

With those economics, it would be hard for Apple (or anyone else) to compete with SpaceX in this business. All else being equal, their cost to create the satellite network would be substantially bigger which means they would have to charge more or have much lower margins and profits than SpaceX.

Right. This seems a lot like a strategy to use/monetize spare launch capacity.

Remember though that Blue Origin is a completely separate corporation from Amazon. That makes me wonder if a Amazon-Blue Origin tie up would be consider self-dealing which is strongly frowned upon.

I don't think the frequencies these satellite ISPs are using is conducive to directly interacting with a phone. The antenna/receiver unit for Starlink will be the size of a pizza box according to SpaceX. Physics limits how much you can shrink the antenna (due to the wavelength).

I thought sats in LEO have fundamental issue that they can’t stay up too long. In rather short time, their fuel would run out and they will burn off in atmosphere.

Does this mean you have to keep replishining them at regular intervals like less than a year or two?

Pretty much - I think Starlink satellites will have an estimated lifespan of 5 years, not bad.

This used to be a huge issue when satellites had to be very big and expensive to launch. But today satellites can be much smaller (several hundred kg) and are much cheaper to launch, making this approach viable.

It’s a “feature” — things self dispose and you don’t end up with lots of junk in orbit

You can see how this is very attractive if you have a lot of launch capacity to sell :)

Do we have the capability to manage the traffic with these 3k Amazon satellites + Starlink's 12k + OneWeb's 650 + all the stuff that's already up there?

Starlink is the only one that mentions collision avoidance during operations.

OneWeb says plenty about collision avoidance if you look at their FCC filings and press statements. Amazon hasn't really gone into detail for what they're doing, but the FCC won't clear them unless they explain how they'll comply with deorbiting guidelines (and it looks like the FCC expects operators of these megaconstellations to do much more than the bare minimum).

It would be great if Amazon Prime included internet access. One time fee for the dish, works with all their in-home stuff.

Didn't FB try do this and scrapped the project?

That was balloons

Google was (ba)loons, FB was solar powered high altitude drones.

As I recall Facebook tried to commandeer the “internet” for Indian citizens by provided a subset of the real internet.

So... Was the subset just all of Facebook and nothing else?

a few dozen websites: https://en.wikipedia.org/wiki/Internet.org#Available_website...

So we'll be at ~20,000 satellites in orbit with this and Starlink. That's four times as many as are currently in orbit.

Well, if we want to generate the efficiencies of scale that would ever allow us to get off this rock, we've got to start with something. There is a lot of money kicking around not doing a whole lot right now, and building rockets is more likely to get us going in the right direction than acqu-hiring another Uber for Cats whizzbang startup for grossly more than it is worth.

The question is what is the limit of LEO such that it causes problems for 1) astronomy such they even with knowledge of passing objects it still causes disruptions in images and 2) that we begin the slippery slope of trapping ourselves

Would have liked a more substantive article on the subject...

Seriously, let’s spend billions in renewable investment and government lobbying to fix the climate crisis. Wow, this is getting silly. Broadband internet is just not a fundamental human need, a habitable planet is.

Sadly, I think there's no incentive for doing well for the planet to get business value. Maybe when we'll run out of clean drinking water, or land, or clean air, then traditional businesses will invest in R&D and come up with ideas and products to sell you to alleviate the pain.. that's the the goal of a business. what we need is more philantropists. Jeff Bezos is not a philanthropist. Bill Gates on the other hand - more likely..

The goal of a business is approximately whatever the founders decide to make it. The view that there is something inherently exploitative about all corporations is just naive.

Tesla is selling billions worth of cars while doing good for the planet. They have also triggered a technological revolution that will do more than any government or the UN could ever do with regulations.

If Bezos successfully ushers in an era where toxic industrial work is moved to space that could do more for the planet than almost anything imaginable. And that is his goal with Blue Origin.

These are the good actors. The bad actors are the Koch Brothers, coal plants, Walmart, the old car companies, etc.

Bringing several Mbps internet to the farthest most unreachable corners of the earth, economically enough for low income people to use, is a net social good. you would understand if you had ever worked in geostationary satellite based internet access and seen how costly it is.

I'm not disagreeing with that, but let's talk about priorities.

If you're living in some of the farthest most unreachable corners of the earth, you;re life is probably more and more in-danger from the threats of climate change through drought, fires etc.

Yet another plan to launch thousands of satellites into orbit so it's easier to order more junk from Amazon is a low priority for those people I'm afraid.

Having worked in developing nation telecom - you might be surprised. Look at the % of people with HSPA+ or LTE data on their $60-80 (in local currency equivalent) Android phones in the eight largest cities in Afghanistan. Look at the growth of developing-nation specialist mobile phone carriers such as south africa's MTN expanding into rural areas of underdeveloped African countries.

It's not like the satellites are going to serve a walled garden Amazon internet, or that their main purpose will be to go to amazon.com and buy products. I'd be very surprised if the service didn't resemble a normal ISP in terms of default route out to the Internet for a full routing table.

Commodity residential broadband makes telecommuting possible. Reducing the influence of population density on capacity benefits folks like me that try to telecommute from rural areas.

I'd be really surprised if the most polluting countries in the world don't have access to broadband?

It doesn't do any good to pick an unrelated endeavor and say the resources would be better spent on a worthier cause instead. Private spending on space, or public funding of arts and culture, or humanitarian aid, etc. is not blocking us from tackling global warming.

If anything, such tangential criticisms are a distraction. Advocate for measures that would help, such as a carbon price and dividend, instead of attacking other causes and making unnecessary enemies.

> If anything, such tangential criticisms are a distraction. Advocate for measures that would help, such as putting a carbon price and dividend, instead of attacking other causes and making unnecessary enemies.

Amazon is allegedly the biggest company in the world, and is still not publishing open carbon emissions data[1]. Is Amazon here an "unnecessary" enemy?

[1] https://www.seattletimes.com/business/amazon/amazon-reluctan...

That has nothing to do with plans for satellites. Especially not space or broadband in general, as the original comment decried.

Feel free to criticize other matters, but don't lump in separate projects.

It doesn't do any good to pick an unrelated endeavor and say the resources would be better spent on a worthier cause instead.

In theory that all sounds fine, however; Wouldn't you say it's time to set some priorities?

“It’s multiple billions of dollars of capex. Amazon is a large enough company now that we need to be doing things that, if they work, can actually move the needle."

Sure, advocating and not criticizing sounds good; Although, what madness and what a waste of time and money. It sounds like all we can hope for is Amazon introduces a self-imposed carbon tax.

If you think this is anything other than post-capitalist market capture disguised as philanthropy, I have some bad news for you.

What's "post-" about it?

Capitalism is free markets, but that quickly devolves into regulatory capture and such, as per Adam Smith, Karl Marx, et. al.

If only!

How is this disguised as philanthropy? He literally is quoted as "a very good business for Amazon"

Reply on news.ycombinator.com