Autonomous Car Progress

[diplomat33]

So any car should be able to operate in any city with Waymo driver since the updates are pushed to all cars no matter where they operate? You could take a car from San Francisco and have it drive to L A and drive around L A once L A is mapped?

Yes, Waymo Driver will drive in any city that is mapped. But Waymo needs to make sure the safety is high enough before they actually passengers ride in the driverless cars. That is why Waymo does testing in each city before they launch driverless ride-hailing.

 

[nativewolf]

So Waymo is testing in 25 metro regions, a few more than I'd have guessed.

Metropolitan statistical area - Wikipedia In the USA NYC, LA, and Chicago are the biggest markets for Uber.

How fast one is scaling in Robotaxi probably could use some definition. Does one look at the number of vehicles, the number of potential clients (addressable market), the amount of revenue (which will be very hard to guess at with most of them).

I was thinking that it would be a combination of those, probably track several factors.

 

[Bladerskb]

Since there is so much angst over easily quantifiable terms such as fast and slow, I encourage you all to do so before continuing this argument. What is "slow" and what is "fast" in terms of scaling? Be precise. Your quantifiable definition will be used later to support, refute, or mock your arguments.

If they did that, how would they be able to continue gainsaying in the future?

Here's a poll to put things into perspective. This is why @Knightshade can say things like below and when called out, responds with "It was true then".
Aka, no matter what happens, he will consider himself right. Trying to have a discourse with him is a waste of time.

the entire point is Waymos methods can't scale.

 

[powertoold]

Waymo will never cover the entire top 50 US cities by population for the general public; my thoughts anyway.

Scaling is not only a matter of technology or approach, it's also a business resource limitation.

You don't want to hear marketing about Waymo tackling this intersection or scenario. You want to hear about Waymo spending considerably less than an Uber per mile. You want to hear about Waymo's costs trending down to below the cost of running an Uber, etc.

 

[diplomat33]

Cruise expands driverless to Austin and Phoenix.

https://twitter.com/x/status/1605233663956127746

 

[Bitdepth]

Waymo will never cover the entire top 50 US cities by population for the general public; my thoughts anyway.

Why would they need to cover the top 50 cities? People just make up arbitrary numbers as if they mean anything. The top 10 cities have a combined 20+ million people in it. There are several companies vying for the market and between them they can cover various cities, granted there will be some competition in certain cities but they don't all need to be in the same cities.

 

[Bitdepth]

Cruise expands driverless to Austin and Phoenix.

https://twitter.com/x/status/1605233663956127746

Well looks like they are hitting their timeline they set.

 

[diplomat33]

Well looks like they are hitting their timeline they set.

Yes they are.

Kyle also tweeted this.

https://twitter.com/x/status/1605234935866220544

I am actually not too surprised that they are seeing improvements from the AVs in Phoenix trickle down to the AVs in SF since it is generalized FSD. That's the same thing Waymo has noticed too. It's one reason Waymo is starting to scale much quicker now since the experience is cumulative and the autonomous driving is getting better everywhere.

 

[Bladerskb]

Waymo will never cover the entire top 50 US cities by population for the general public; my thoughts anyway.

Scaling is not only a matter of technology or approach, it's also a business resource limitation.

You don't want to hear marketing about Waymo tackling this intersection or scenario. You want to hear about Waymo spending considerably less than an Uber per mile. You want to hear about Waymo's costs trending down to below the cost of running an Uber, etc.

Waymo will cover 20-25 cities / distant areas out of the top 100 Uber cities by earnings in the US by the end of 2025 and 40-50 cities / distant areas by end of 2028.

2 Distant areas = (100 sq.mile downtown Phoenix and 50 sq mile Easy Valley suburban Phoenix)

32 cites and areas Uber and Lyft drivers earned the most in December

The Bay Area in California takes first place, with drivers earning on average $23.28 per hour in December 2019.

www.businessinsider.com

 

[diplomat33]

This is why Waymo and Cruise are scaling faster than the naysayers thought. It won't take years to scale to each city.

https://twitter.com/x/status/1605234928916267008

 

[ItsNotAboutTheMoney]

For their model:
(1) Reach profitability
(2) Start rental company that doesn't suck.

 

[diplomat33]

Here are the Cruise geofences for the general public driverless rides. They are small for now. But it is a start. Cruise has a footprint now in all 3 cities. They will expand the ODD.

 

[stopcrazypp]

Why would they need to cover the top 50 cities? People just make up arbitrary numbers as if they mean anything. The top 10 cities have a combined 20+ million people in it. There are several companies vying for the market and between them they can cover various cities, granted there will be some competition in certain cities but they don't all need to be in the same cities.

They don't "need" to, but it would be a relatively meaningless technology for many geographic areas if it only covers the top 10 cities. Uber for example scaled much faster. If people never see a robotaxi in their city in their lifetimes, then it's as good as worthless as a technology for them (and all the talk about the next generation not needing to learn to drive would not come true).

 

[Bitdepth]

They don't "need" to, but it would be a relatively meaningless technology for many geographic areas if it only covers the top 10 cities. Uber for example scaled much faster. If people never see a robotaxi in their city in their lifetimes, then it's as good as worthless as a technology for them (and all the talk about the next generation not needing to learn to drive would not come true).

Did you miss the part where i said there are various companies vying to provide these services and they don't all need to be in the same cities? Between them they can cover various markets without all of them having to be in the same cities.

 

[diplomat33]

They don't "need" to, but it would be a relatively meaningless technology for many geographic areas if it only covers the top 10 cities. Uber for example scaled much faster. If people never see a robotaxi in their city in their lifetimes, then it's as good as worthless as a technology for them (and all the talk about the next generation not needing to learn to drive would not come true).

A couple things.

1) Uber scaled much faster but it was a completely different type of scaling problem. Uber did not need to invent a new tech like autonomous driving. Uber basically just needed to design a ride-hailing app and scale the number of drivers. There were logistical and economical problems to solve but in the case of Waymo or Cruise they need to solve a new tech AND solve the logistical and economical problems too. So scaling is a much harder problem for Waymo to solve than for Uber. That is why I think that it is simplistic to compare the two.

2) I do agree that simply scaling robotaxis to 10 cities is not good enough. The true goal is to scale AVs everywhere, so everybody has access to it. The fact is that 1 company is not going to scale AVs to every US city. That is unrealistic. We can expect different AV companies to divide up the market share. So do I expect that between Waymo, Cruise, Zoox, Motional, Mobileye and maybe a few others, the combined companies will eventually cover most major US cities.

I also think that personal cars is the best way to really scale AV tech to the most people. I see robotaxis as a means to an end because it is a great way to test and deploy the tech. But when AV companies are able to deploy L4 to affordable consumer cars, that is when the tech will really scale to everybody. so the real question is when will Waymo, Cruise, Mobileye and others put their L4 on consumer cars. Now, I am sure people will point out that Waymo has not said anything about wanting to put their L4 on consumer cars. True. But I personally, think they will once the tech is ready. The L4 needs to be a big enough ODD. Also, Waymo needs to be confident the safety is good enough everywhere before they would put their L4 on consumer cars. But once the tech is there, I think Waymo would be silly not to do some sort of L4 on personal cars since it would be a great way to spread the tech and generate revenue. And Waymo says their goal is to get their tech to as many people as possible. If that is their goal then they are not going to just deploy robotaxis in a few US cities and call it a day.

 

[diplomat33]

https://twitter.com/x/status/1605274918693883929

 

[stopcrazypp]

Did you miss the part where i said there are various companies vying to provide these services and they don't all need to be in the same cities? Between them they can cover various markets without all of them having to be in the same cities.

Of those various companies, none are anywhere as close as Waymo and Cruise (and some have given up). It's not like this is a fully open source tech (like for example Android for phones), where there can be third party players that can directly use the same tech.

So I think it is unrealistic to expect the various companies to step up and cover other smaller areas if even the biggest ones decide they are not worthwhile. Edit: Note I'm talking about fleet based companies as below. For companies focused on personal vehicles, theoretically one company can cover the entire country (there is no need for splitting the market up).

 

[stopcrazypp]

A couple things.

1) Uber scaled much faster but it was a completely different type of scaling problem. Uber did not need to invent a new tech like autonomous driving. Uber basically just needed to design a ride-hailing app and scale the number of drivers. There were logistical and economical problems to solve but in the case of Waymo or Cruise they need to solve a new tech AND solve the logistical and economical problems too. So scaling is a much harder problem for Waymo to solve than for Uber. That is why I think that it is simplistic to compare the two.

2) I do agree that simply scaling robotaxis to 10 cities is not good enough. The true goal is to scale AVs everywhere, so everybody has access to it. The fact is that 1 company is not going to scale AVs to every US city. That is unrealistic. We can expect different AV companies to divide up the market share. So do I expect that between Waymo, Cruise, Zoox, Motional, Mobileye and maybe a few others, the combined companies will eventually cover most major US cities.

I also think that personal cars is the best way to really scale AV tech to the most people. I see robotaxis as a means to an end because it is a great way to test and deploy the tech. But when AV companies are able to deploy L4 to affordable consumer cars, that is when the tech will really scale to everybody. so the real question is when will Waymo, Cruise, Mobileye and others put their L4 on consumer cars. Now, I am sure people will point out that Waymo has not said anything about wanting to put their L4 on consumer cars. True. But I personally, think they will once the tech is ready. The L4 needs to be a big enough ODD. Also, Waymo needs to be confident the safety is good enough everywhere before they would put their L4 on consumer cars. But once the tech is there, I think Waymo would be silly not to do some sort of L4 on personal cars since it would be a great way to spread the tech and generate revenue. And Waymo says their goal is to get their tech to as many people as possible. If that is their goal then they are not going to just deploy robotaxis in a few US cities and call it a day.

Your last point is what I was getting at. If all the companies stay as fleet based robotaxi companies, I see it as unrealistic to expect various companies to spread out geographically to cover all areas (or anywhere close), especially if they can't even match Uber's pace. For the self driving future to really come to the masses in the less populated areas, it'll be up to personal vehicles, not the fleets.

 

[Bladerskb]

This is why Waymo and Cruise are scaling faster than the naysayers thought. It won't take years to scale to each city.

https://twitter.com/x/status/1605234928916267008

But but @Knightshade told us it takes Waymo (and Waymo-like approaches aka Cruise) 5 years to HD map a city.
Ofcourse he will respond and tell us how he was "right".

How many offer RT service to the public after years of mapping work? Zero.
If it takes you 5+ years to get any given city mapped and running, that's not scalable.
If it were "automated" it wouldn't take years to do.
Waymo appears to still take years to get ONE city mapped...

See again the fact they've been "working on" mapping a bunch of cities for years and rolled out service to none of them (except the very limited internal SF testing that appears to be happening after like 5 years of mapping work)

 

[Knightshade]

But but @Knightshade told us it takes Waymo (and Waymo-like approaches aka Cruise) 5 years to HD map a city.
Ofcourse he will respond and tell us how he was "right".

I mean, the guy you quoted doesn't work for Waymo, so you again appear to be spamming the forum with non-sequiturs.

Waymo, as a reminder, began testing in SF in 2009


To be fair to Waymo though- the post of mine you quoted pointed out they began mapping in LA in 2019... and it appears they want to offer service there in 2023.

So that's only FOUR years, not FIVE between mapping and offering service. QUITE AN IMPROVEMENT!

Assuming they actually deliver on time of course.