Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
   Glossary   Glossary
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Pages: 1 [2]

Author Topic: Why Starlink is doomed to fail  (Read 1588 times)

celso

  • Member
  • **
  • Posts: 34
Re: Why Starlink is doomed to fail
« Reply #15 on: June 13, 2022, 02:31:46 AM »

Most people wont have a clue how Autopilot on a plane works whereas pilots will have had extensive training on EXACTLY how to use it, what it does and does not do.  Someone buying a car will not have the same knowledge, nor do you have the same amount of time to react to a problem when you're on a road compared to a plane.

Even if it tells you exactly what it does and does not do in the manual, it has to be to a higher standard as you cannot expect someone who is just making a trip to the shops, potentially with many distractions, to be aware of everything it does and does not do at all times.  I stand by my claim, its dangerous for ANY car to "partly" self drive, as it causes complacency by the driver if most of the time they do not have to steer.  At most it should be able to emergency brake, park, stay in lane, maintain speed on a motorway.  The driver should ALWAYS be the one changing lanes, turning corners, etc.  We have decades of proof that with the best intentions, people become complacent once a majority of the time they don't have to do something, making them slow to react when they are required to.  By its nature, by the time you realise its gone wrong, you may be too late to react.

Its risky to even expect a car to emergency brake:
watch?v=p7lp5f0aqzU
Had this been a real-world scenario with traffic coming in the other direction, things could have gone badly.

My point being, you shouldn't be promoting a system as able to do something unless it can do it 100% of the time.  The driver shouldn't just have their hands on the wheel, they should be steering, braking, etc, you should NEVER be relying on the car to react period.  But this is not how Tesla are promoting their tech however.

Autopilot is cruise control+line keeping.

I understand if non-Tesla owners don't know the difference between this and full self driving, but I find it hard to believe that owners don't have a basic idea of what the system does:

1. When ordering a car, you are told the differences between Autopilot and the other paid options. You can't skip the page.
2. Autopilot needs to be enabled manually and the displays tell you what you should do (be ready to take over, pay attention).
3. If you remove your hands from the steering wheel, after a while you'll see and hear warnings to hold the wheel. Newer cars have a camera pointed at the driver too.
4. As you use the feature, you understand and experience its limitations.

Most Tesla owners are informed by point 1, 2, and 3. For those renting, leasing, and buying used cars, there's still points 2 and 3 (no one escapes this).

The system itself isn't special. It maintains speed, distance, and keeps the car inside the lines. Compared to others, it's actually good. My father owns a 2016 Toyota and there are buttons in the steering wheel to set cruise control, increase/decrease speed, etc... but the car doesn't slow down, keep a certain distance, steers, etc. If you don't brake yourself, you'll hit whatever is in front of you.

It's a valid opinion to think that such incomplete systems shouldn't allowed on the road, but if some youtuber uses this as a "prof" that Tesla (or any other brand) is bad and a scam, I just can't take them seriously.

The reason I put part of the blame on the user here is simple. You are the person in control and you *have* to correct the car's mistakes. There have been accidents with Autopilot where the driver was seeing the problem for a few seconds (eg: a truck crossing the road ahead) and they did nothing. You have this huge thing pulling in front of you and... no reaction? Some people will adapt to these systems better than others, but if you're not good with it, why would you keep using it?

And then you have those who work around the system to be on the phone or something like that... this is where I have zero tolerance. It's a "people that shouldn't be allowed on the road" problem, not a problem with the car systems.

My opinion would be completely different if Autopilot was self driving, but it isn't. You need hardware, an account with good score, join the beta, see a few more warnings to use that. If the car tells the driver that it can't self drive and the driver ignores everything and then crashes, I find it hard to blame the car.

Anyway, clearly the "autopilot" name is a problem, so much so they're being investigated in the US and I just read a news article how they may have to be forced to "recall" the software. They should just call it "adaptable cruise control with line assist". I'm sure that would end the "confusion".

I've watched that video, he's using autopilot. Is it expected for cruise control to bring the car to a stop or go to the other lane to avoid a crash?

I now his videos because there was a discussion a few months ago over at Hacker News about him. He worked for Tesla and made videos of his own car using the self drive beta. It didn't look good (the software is clearly not ready!), but he also would let the car do the mistake instead of taking over. He was fired from Tesla and access to the beta was removed from his account.

Firing him doesn't look good for Tesla. If the beta software is so good, why hide it? On the other hand, it also doesn't look good to have an employer letting the car hit things on a public road. The "tesla employer shows how bad their self drive is" headlines probably didn't help.

I can't find the video right now, but this again comes down to economics.  The weight of the batteries required for a truck means the trucks can haul a tiny minority of what a diesel truck can.  Musk made claims about cost per distance, but didn't factor this in, making it appear cheaper and more efficient when its not.

In fact, a lot of the problems come down to energy storage.  We've pretty much peaked with battery storage, we can't make it much more efficient, the laws of physics dictate so.  The only way to make electric trucks work is this:
watch?v=_3P_S7pL7Yg

I'll wait to see what they'll do. I remember reading something about them renting the truck instead of selling them? I don't think anyone was scammed. Worst case there's no truck or it's not a viable product and anyone thinking of renting can use something else or pull out of any deals (I believe that's how it usually goes).

The factory where the Semi is going to be manufactured was completed in Q1 (from the info I found online, could be wrong). Their information for investors says 2023. We'll see what happens soon, I guess.

The idea from that Tom Scott video is interesting, in some places they already use the idea for buses, but I don't know if countries will go with different solutions when it's easier (just the amount of complaints about how the wires look...) to have chargers and more frequent stops. I don't know what's the solution most countries will pick.

Inconsistent range due to improper charging, that's not a small issue especially for long distance travel in rural areas where you could get stuck in the middle of nowhere.

watch?v=eSOHzmqLWjU

I wasn't aware. If that's the case, then they should fix the faulty hardware. No excuses or bs fixes.

This is not a "what about X" excuse, but since some of the videos in this thread are about tesla/spacex being scams, we should keep in mind that this is the expected behaviour from most car manufacturers (one of the reasons why we shouldn't be "loyal" to a brand). For example, the Chevrolet Bolt/Opel Ampera-e received a few software updates before they finally replaced all batteries for free.

As for Viasats negatives, the problem is you're looking at it from a "this is what we want" scenario, rather than "this is what we can do".

We all want perfect broadband that works for everything, but if its not economical or damaging the environment, then we absolutely should NOT be doing that.  The end does not justify the means.

Its the same kind of logic as "we need land for x, lets just cut down all the trees and kill off all the wildlife".  Sure, it will wipe out all life on earth eventually, but things will be fine for a short while.  Are you honestly advocating that we should just ignore all long-term problems in order for a quick fix today?

This is the crux of why I'm critical of Elon Musk, even though I wont be using any of his technology.  Because it WILL impact everyone in the long term.

I'm with you in regard to the negative side effects of so many satellites. I also don't see myself using any satellite based internet, so my "defence" of it isn't because I want to have it... but we need to be realistic. Other space alternatives aren't as good and like it or not, this is happening.

There are many places without internet/ground networks or with a really bad service, so there's demand from regular people for good space based internet. And then you have things like the military, network providers looking for "backhaul" when land based isn't available/is more expensive, companies wanting fast and cheap internet in their ships and planes, etc. The demand is there and if company/country A doesn't do it, company/country B will.

I don't expect everyone to reach a compromise and use just one constellation because the countries that can do this don't trust each other. That's why we have GPS, Galileo, Glonass, Beidu, etc, even though they do the same thing. I guess someone could ban these constellations, but good luck trying to get everyone to agree or to enforce the ban.

Musk could cancel Starlink tomorrow and de-orbit all satellites. Instead of 6 or 7 by the end of the decade, we'd have one less network. I'd rather start working on something to manage the orbits of all these satellites because just like the trees being cut down, those satellites will be up there.

About Musk himself, I don't get the "cult" of loving or hating the guy and he sometimes seems to be a bit of a tw*t, but that doesn't make Starlink more or less viable. What "trigged" my reply here was the speeds and comparison with fibre(!) and not accounting for other sources of revenue. I'm probably wrong about some things, but that video has its problems :P

Anyway, I don't think I can add much more to this. Time will tell if his companies and respective products/services are good or not.
« Last Edit: June 13, 2022, 04:12:37 PM by celso »
Logged

digbey

  • Member
  • **
  • Posts: 56
Re: Why Starlink is doomed to fail
« Reply #16 on: June 13, 2022, 12:16:12 PM »


The idea from that Tom Scott video is interesting, in some places they already use the idea for buses, but I don't know if countries will go with different solutions when it's easier (just the amount of complaints about how the wires look...) to have chargers and more frequent stops. I don't know what's the solution most countries will pick.

Looks like they've reinvented the trolleybus. Single deck and double decker buses like this were operating in many cities in the UK from 1911 to 1972 and are still in use in some European cities.

https://en.wikipedia.org/wiki/Trolleybus

In the 1950's I used to go to school in a London Transport double deck trolleybus, it was always preferable to the noisy diesels.
« Last Edit: June 13, 2022, 12:18:56 PM by digbey »
Logged

celso

  • Member
  • **
  • Posts: 34
Re: Why Starlink is doomed to fail
« Reply #17 on: June 13, 2022, 09:18:21 PM »

Looks like they've reinvented the trolleybus. Single deck and double decker buses like this were operating in many cities in the UK from 1911 to 1972 and are still in use in some European cities.

https://en.wikipedia.org/wiki/Trolleybus

In the 1950's I used to go to school in a London Transport double deck trolleybus, it was always preferable to the noisy diesels.

I didn't know London used to have those. When I moved here ~15 years ago, every bus I used was noisy. Then they introduced hybrids.

At least on some of them the engine is only there to generate electricity for the electric motors/battery. Often the bus starts moving quietly and then you can hear (and feel) when the engine kicking in (apparently this still more efficient than using a diesel engine powering everything).

Now there are some routes with fully electric buses. They're quieter. The air conditioning still makes noise though (first world problems... :lol: ):

https://www.youtube.com/watch?v=e13m-vOU2mU

I wonder if a system with pads on the ground at main bus stops could. The people behind this system claim good results (no idea if it's actually the case or not):

https://www.youtube.com/watch?v=aq7SP18sPKw
Logged

Alex Atkin UK

  • Kitizen
  • ****
  • Posts: 4120
    • Thinkbroadband Quality Monitors
Re: Why Starlink is doomed to fail
« Reply #18 on: June 14, 2022, 07:37:37 AM »

Yeah hybrid is great, as I understand it you use a ton of fuel accelerating so if you do that with electric instead from a battery (which can be trickle-charged when driving), then I can see why that would be far more efficient.   I've been in a few Ubers using hybrids too, they're a much nicer, quieter ride.  Though in a dense city environment, I can see how fully electric can work for things like buses too, especially if they can figure out how to charge when sat at the terminus.

Unfortunately I do not believe induction pads are an option. They're insanely inefficient so you're wasting at best 2/3 of the power consumed, I'd suspect a lot more given the huge gap between the coils.  The same reason so many wireless power companies have completely failed.

It works for charging phones as the coils are mm apart, and even there its about 40% efficient.  It works great for cooking, as you WANT to create heat, and that's where that other 60% is going.
Logged
Broadband: Zen Full Fibre 900 + Three 5G Routers: pfSense (Celeron N5105) + CPE Pro 2 H122-373 WiFi: Zyxel NWA210AX
My Broadband History & Ping Quality Monitors

celso

  • Member
  • **
  • Posts: 34
Re: Why Starlink is doomed to fail
« Reply #19 on: June 14, 2022, 01:03:12 PM »

I don't have the equipment to test, but I've read that the latest iPhone using Apple's charger can get up to ~75%. More efficient components cost more though. I doubt my OnePlus 8 Pro with the cheap wireless charger I have (cheap one from Aliexpress) can be that good. I should test and compare it to the cable I normally use.

In any case, I don't think wireless will ever beat a wired solution, but I'm not seeing them installing those overhead cables in cities these days... a cell tower is what it is, imagine cables and everything needed to support them.

Maybe something like this could work? Use batteries and a capacitor, then charge the capacitor in places you know the bus/truck will stop/pass?

https://en.wikipedia.org/wiki/Capacitor_electric_vehicle
Logged

Bowdon

  • Content Team
  • Kitizen
  • *
  • Posts: 2354
Re: Why Starlink is doomed to fail
« Reply #20 on: June 14, 2022, 06:54:49 PM »

I've not watched the video in the OP. But I've never thought things like Starlink and other similar businesses will be able to compete along side (the new) traditional methods like full fibre and the 5/4G technologies.

I think it would be a good backup, or to cover areas that conventional systems struggle for a financial case to cover.

But I think it'll end up having to drop its price as it'll be a third class system (behind fibre/cable and mobile broadband) technology.
Logged
BT Full Fibre 500 - Smart Hub 2

Alex Atkin UK

  • Kitizen
  • ****
  • Posts: 4120
    • Thinkbroadband Quality Monitors
Re: Why Starlink is doomed to fail
« Reply #21 on: June 14, 2022, 08:29:37 PM »

But I think it'll end up having to drop its price as it'll be a third class system (behind fibre/cable and mobile broadband) technology.

The crux of the matter is at its CURRENT prices, it runs at such a large loss that it can never be profitable.

The old dish cost them $2000 to make and they only charged a $500 installation fee, I doubt the new one is so dramatically cheaper that its now profitable.
.
More importantly, it costs $67 million to launch a Falcon 9 rocket.  Even if that's carrying 143 sattelites, if you do the maths it simply doesn't add up.  The last launch only carried 53, and every one of those satellites only has a 5 year expected lifespan.

Maybe something like this could work? Use batteries and a capacitor, then charge the capacitor in places you know the bus/truck will stop/pass?

https://en.wikipedia.org/wiki/Capacitor_electric_vehicle

Its an interesting idea, but the infrastructure would be a nightmare as every time the bus stops it would put a huge load on the electricity grid.

Also having seen small capacitors explode, I'm not sure I like the idea of sitting on a super-capacitor.  :-\
« Last Edit: June 14, 2022, 08:39:12 PM by Alex Atkin UK »
Logged
Broadband: Zen Full Fibre 900 + Three 5G Routers: pfSense (Celeron N5105) + CPE Pro 2 H122-373 WiFi: Zyxel NWA210AX
My Broadband History & Ping Quality Monitors

celso

  • Member
  • **
  • Posts: 34
Re: Why Starlink is doomed to fail
« Reply #22 on: June 14, 2022, 11:59:34 PM »

SpaceX sells a Falcon 9 launch to 3rd parties for $67 million and offers "modest discounts" for multiple launches (March 2022):

https://www.spacex.com/media/Capabilities&Services.pdf

Even if one ignores the rumours about it costing them half this amount to launch and the fact they re-use the same booster many times for Starlink launches, it makes no sense to assume they make no money on launches and that they even make a loss on customers that launch more than once.

According to this article from November 2021, SpaceX says the original cost of the dish was $3000. In April 2021 it was around $1500 (according to the CNBC article this page links to) and was $1300 when this article was posted in November. They quote Gwynne Shotwell saying that "The ones we will have later this year will cost roughly half of what our current user terminals cost". Half of $1300 is $650 (edit: or $750, if we use the $1500 from April).

In March 2022 they increased the price of their kit from $499 to $599.

The whole thing may never be profitable, but you're using inflated numbers Alex.
« Last Edit: June 15, 2022, 02:24:03 AM by celso »
Logged

kitz

  • Administrator
  • Senior Kitizen
  • *
  • Posts: 33212
  • Trinity: Most guys do.
    • http://www.kitz.co.uk
Re: Why Starlink is doomed to fail
« Reply #23 on: June 15, 2022, 12:32:02 AM »

I'm unable to confirm those figures which seem astronomical.  I'm also sceptical of YouTube sceptic type articles. "Debunking Starlink" by 'Common Sense Skeptic'" alerts me to the type of audience the author is trying to appeal to.... and that the topic is likely to have an axe to grind or full of conspiracies.   
That said, the author name and article title ignored, he does raise some valid concerns. Even without checking figures, the project seems unviable purely because it relies on the assumption that such as large number of subscribers would be needed to return a profit... and more importantly its not bringing anything new to the table that would actually advance broadband technology.  When the rest of the world is trying to expand FTTH networks consumers will not be interested in a slower product with extremely high latency costing more than double. 

One of the things I have noticed is that a lot of Americans dont understand the technology - Why should they?   Just like here, many of them have no idea of the difference between satellite, wifi and cell.  Another thing I've noticed is outside of the major cities, there doesnt seem to be many major advancement plans to roll out FTTH.   There is no equivalent of Openreach roll-out.   Many towns and even cities have ISPs who are localised.  I have a friend in Albuquerque and I'm truly amazed at limitations for broadband.  Whilst city centre may have FTTC, just 10 miles from the city center choice is very limited.  My friend has recently moved nearer to the University of NewMexico  where she works, but previously just 15mins from work she was limited to satellite.   The ISP was basically a one man band and her connection used to be notoriously slow. Hard to believe when its not like she lived in the middle of nowhere.

Was going to type more, but am going to have to cease soon...   questions such as why the FCC as a regulator isnt doing more when it comes to gov funding.  It should know what is feasible and what isnt. 

As regards to the super highway trucks, I was reminded of railroad transportation. Somehow I cant imagine one lane working like that for the M25. 
I even have doubts if electric & batteries really are a more efficient form of power... especially when fossil fuels are needed to charge the batteries. 
Logged
Please do not PM me with queries for broadband help as I may not be able to respond.
-----
How to get your router line stats :: ADSL Exchange Checker

Alex Atkin UK

  • Kitizen
  • ****
  • Posts: 4120
    • Thinkbroadband Quality Monitors
Re: Why Starlink is doomed to fail
« Reply #24 on: June 15, 2022, 02:53:47 AM »

Yeah I have a friend in Texas, his parents only got FTTC last year.  Although I think he might have access to Cable but despises the company.  Some Americans seem to have some very odd loyalty to telcos, even when they HAVE options.

Not entirely sure what speed him or his parents have as he doesn't consistently max out my upload, but transferring data to Texas bottlenecks even to a VPS.

Electric vehicles is working on the assumption at some point we'll only have nuclear, wind and solar, with some kind of storage.  Not sure its feasible, especially anything we'd have to import from abroad.

Its also unclear if solar is remotely clean, given making the panels uses a lot of energy and they're toxic so problematic if they do not get recycled at the end of life.  You don't want solar panels in landfill, then again half the things people put in the household waste probably shouldn't be either.

SpaceX sells a Falcon 9 launch to 3rd parties for $67 million and offers "modest discounts" for multiple launches (March 2022):

https://www.spacex.com/media/Capabilities&Services.pdf

Even if one ignores the rumours about it costing them half this amount to launch and the fact they re-use the same booster many times for Starlink launches, it makes no sense to assume they make no money on launches and that they even make a loss on customers that launch more than once.

According to this article from November 2021, SpaceX says the original cost of the dish was $3000. In April 2021 it was around $1500 (according to the CNBC article this page links to) and was $1300 when this article was posted in November. They quote Gwynne Shotwell saying that "The ones we will have later this year will cost roughly half of what our current user terminals cost". Half of $1300 is $650 (edit: or $750, if we use the $1500 from April).

In March 2022 they increased the price of their kit from $499 to $599.

The whole thing may never be profitable, but you're using inflated numbers Alex.

I think this says it all really https://www.nasdaq.com/articles/elon-musk-predicted-starlink-would-generate-%2430-billion-in-revenue.-dont-believe-it.-2021
« Last Edit: June 15, 2022, 03:22:22 AM by Alex Atkin UK »
Logged
Broadband: Zen Full Fibre 900 + Three 5G Routers: pfSense (Celeron N5105) + CPE Pro 2 H122-373 WiFi: Zyxel NWA210AX
My Broadband History & Ping Quality Monitors

celso

  • Member
  • **
  • Posts: 34
Re: Why Starlink is doomed to fail
« Reply #25 on: June 15, 2022, 05:23:56 AM »

I'm sorry, it's not my intention to get personal, but you are ignoring information that doesn't fit your view and one simply can't reach proper conclusions by doing that.

- It doesn't cost SpaceX 67 million to launch their own rocket with their own payload. If it's half of the cost, they can launch 2x more rockets for the same price. You can't ignore this when it's a big part of Starlink costs.

- The price of their terminal is indeed a problem and you are correct to point it out. But they seem to have that mostly under control, yet you're using the $2000 number from more than a year ago. Breaking even or subsidising 100-200 dollars isn't the same as giving 1500 dollars away.

- Previously, you and the youtuber failed to account for the revenue from other sources, and anyone that pays attention to military/financial contracts knows that they will not pay 99 dollars for the service. Not to mention the assumption that this is competing with fibre users.

So you over estimate the cost of launch, how much they're losing with the hardware, and ignore what is likely to be a substantial part of their revenue. Don't you see a problem with this?

Now you linked to that early 2021 article. Musk is wrong about a lot of stuff, so he's probably going to miss that 30 billion target. But again, you're supporting your position by using an article that uses outdated information to make an assessment on the future of Starlink. They're assuming the satellites/capacity will remain the same and that they'll be launched on Falcon 9! This is not true!

I'll quote part of this article: https://www.zdnet.com/article/elon-musk-says-starship-needs-work-to-send-spacex-starlink-2-0-satellites-into-orbit/

Quote
Elon Musk has admitting he is banking on Starship, a launch rocket currently in development, to get SpaceX's next generation Starlink satellites into orbit.

"We need Starship to work and to fly frequently, or Starlink 2.0 will be stuck on the ground," the tech billionaire told YouTube show Everyday Astronaut.

He explained that sending Starlink 2.0 into orbit with the company's Falcon 9 rockets, which were used to send the first generation of Starlink satellites into space, is not plausible.

"Falcon has neither the volume nor the mass to orbit capability required for Starlink 2.0. Even if we shrunk the satellite down, the total up mass of Falcon is not nearly enough to do Starlink 2.0," Musk said.

According to Musk, the first Starlink 2.0 has already been produced, weighing about 1.25 tonnes and measures about seven metres long. He also believes Starlink 2.0 will supersede by an order of magnitude SpaceX's first generation of Starlink satellites.

This is not new information. The new satellites have more capacity. They're not going to use Falcon 9 to launch most of their satellites.

If after being presented by new information, backed with sources, you still insist on using outdated information or ignore some stuff, then all replies here have been a waste of time.

To be clear: Starlink may fail. I don't know if they'll manage to have enough customers and revenue to be successful. The problem for me, in the context of this thread, is that you and that youtuber are using bad information to reach the conclusion that it will fail.
« Last Edit: June 15, 2022, 05:26:08 AM by celso »
Logged

neil

  • Reg Member
  • ***
  • Posts: 338
Re: Why Starlink is doomed to fail
« Reply #26 on: August 12, 2022, 08:42:50 PM »

so far not allowed to operate here
Logged
VDSL FTTC 35/18

Alex Atkin UK

  • Kitizen
  • ****
  • Posts: 4120
    • Thinkbroadband Quality Monitors
Re: Why Starlink is doomed to fail
« Reply #27 on: August 13, 2022, 12:22:42 AM »

The problem is, you are using figures by SpaceX themselves, which are unrealistic.  They're all goals, not based on reality.

Several debunkers have covered how SpaceX claims of revenue for selling space for payloads is insane - there isn't nearly that much demand for launching things into space.  Not to mention the environmental damage from burning all that fuel.

Plus as already pointed out, if Starlink do launch as many satellites as planned, the space junk problem will quickly become such a problem that the whole thing comes tumbling down, including far more essential satellites.  We will no longer be able to launch anything into space ever again.  We wont be able to monitor for meteors, or have GPS, etc.

I fail to see how people in rural locations having lower latency broadband is worth that very high price, even IF the project was economical, which it inherently never will be.

Logged
Broadband: Zen Full Fibre 900 + Three 5G Routers: pfSense (Celeron N5105) + CPE Pro 2 H122-373 WiFi: Zyxel NWA210AX
My Broadband History & Ping Quality Monitors
Pages: 1 [2]