Autonomous Car Progress

[Doggydogworld]

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.

The Austin area is ~5 square miles in the high density downtown areas. High congestion, low speed, similar to San Francisco. The "Phoenix" area is also ~5 square miles, but in Chandler instead of downtown Phoenix. Much easier driving environment than San Francisco, but with 45 mph main roads vs the 25 mph they've historically run. It'll be interesting to see how they handle that.

 

[diplomat33]

Good article from Brad Templeton on the recent Waymo and Cruise expansions:

Waymo And Cruise Make Big Boosts In Robotaxi Service

Cruise and Waymo, the clear #2 and #1 (respectively) in the Robotaxi race, have recently expanded their service areas for public access rides and driving with nobody in the vehicle.

www.forbes.com

Here is a summary of all the recent news:

1. Cruise announced it will begin service in Chandler Arizona (roughly the same area as Waymo’s first service) tonight with service from 7pm to 2 am.
2. Cruise will also start service in a small area around downtown Austin and UT Austin from 10 pm to 5:30 am, the same hours they use in San Francisco.
3. Cruise in November expanded service in San Francisco to all hours and most of the city, except the busiest part of the CBD — but only for use by Cruise employees.
4. Waymo recently expanded service in San Francisco to add the full downtown only for employees and guests. Waymo continues to operate 24 hours/day, but adds safety drivers during certain weather conditions. Waymo service outside the NE quadrant of the city has been available to members of the public.
5. Waymo doubled its service area around downtown Phoenix. Previously it also opened service from there to Phoenix airport, via a stop at the PHX Skytrain people mover that serves all terminals.
6. Waymo has said opening service in Los Angeles is imminent.
7. Waymo is anticipating it will soon get a permit to charge money for rides. The current service, while technically open to the public, has a waiting list.
8. Waymo disclosed it now has 700 robotaxis in operation or testing in is various test territories.

Brad will share about his recent Waymo ride:

"I recently rode in Waymo’s service in the Noe Valley area of San Francisco and the ride was smooth and impressive. There will be more about that ride in a forthcoming article."

He thinks the SF central business district will be a commercially viable service for Waymo:

"Once Waymo allows members of the public access to the central business district, and can charge money, they will have the first commercially viable robotaxi service."

Profitability is still a challenge due to high R&D and logistical costs:

This is not to say that Waymo’s service will be profitable. Indeed, they still have massive R&D and logistical costs that make that far away.

But the full SF service area is a viable taxi market and could convince people to give up some car ownership, but not all:

The full San Francisco service area is both a viable taxi market — after all this is where Sidecar, Lyft and Uber were born — but more importantly a potential market for car replacement. A service established here, with lower prices than Uber, could convince San Franciscans to give up ownership of a car. At present, it would probably be a drop from 2 or 3 cars, not a drop to having no cars, but this is a start.

 

[diplomat33]

Some Waymo rides in downtown Phoenix are seeing some high wait times.

Here we see an example of a ride with a 24 mn wait time:

Here we see a ride with a 19 mn wait time.

The high wait times could mean high demand and/or low supply. If high demand, it is a good problem to have. But I think Waymo needs to bring down the wait time.

I think Waymo needs to increase their fleet size. 700 is not going to be enough to service chandler, downtown Phoenix, SF and soon LA and have vehicles for testing. So it will be interesting to see when Waymo announces a big increase in fleet size. I think that will be a really good sign.

I think the Geely vehicle is still at least a year out. So I imagine they would need to add more I-Paces now before they get the Geely vehicle. But perhaps the Geely vehicle will be ready sooner. It will also be interesting to see how many Geely vehicles they get. That will also be very telling.

On a side note, the cost of the rides seem pretty good. So that is promising that Waymo will be competitive with Uber.

 

[Knightshade]

Dolgov said it does work safely in rain and wet roads.

Also you (bold added):

Waymo continues to operate 24 hours/day, but adds safety drivers during certain weather conditions

LOL as the kids say.

 

[diplomat33]

Also you (bold added):

LOL as the kids say.

There is no contradiction. "working" and "removing the safety driver" are different things. it's a question of reliability. There is a big difference between working and having the data that proves that it works with 99.99999% reliability. The system can work but you can't remove the safety driver until you are sure that it works with 99.99999% reliability. So the Waymo Driver does work in rain. Dolgov showed examples of it working in rain. We also have a video, that I know of, of an entire ride where it works in rain with zero interventions. BUT Waymo wants to make sure that it works with 99.99999% reliability before they permanently remove the safety driver. It's cute that you thought you caught me in a contradiction. I'll give you points for trying but you fail.

 

[Knightshade]

There is no contradiction. "working" and "removing the safety driver" are different things. it's a question of reliability. There is a big difference between working and having the data that proves that it works with 99.99999% reliability. The system can work but you can't remove the safety driver until you are sure that it works with 99.99999% reliability. So the Waymo Driver does work in rain. Dolgov showed examples of it working in rain. We also have a video, that I know of, of an entire ride where it works in rain with zero interventions. BUT Waymo wants to make sure that it works with 99.99999% reliability before they permanently remove the safety driver. It's cute that you thought you caught me in a contradiction. I'll give you points for trying but you fail.

I mean, if we define working as "works in at least one video example with zero interventions" then FSD is also working

 

[diplomat33]

I mean, if we define working as "works in at least one video example with zero interventions" then FSD is also working

Yes, I consider FSD Beta to be working, just not reliable enough to remove driver supervision. I make a distinction between working and working reliably. I also make a distinction between working reliably and having the data that proves that it works reliably. You need the data that proves reliability before you can remove driver supervision. I think I am being consistent in my definitions.

 

[Bladerskb]

Also you (bold added):




LOL as the kids say.

Typical Tesla fans logic.

"If it has a driver it doesn’t count... unless it’s Tesla. Then having a driver behind the system still counts!

- Tesla with a driver that needs to grab the wheel every min to prevent a crash = It Works!
- Waymo with a driver for extra precautions = It doesn’t work! Lidar & HD Map is doomed! Lidar doesn't work in the rain!"

 

[Knightshade]

Typical Tesla fans logic.

"If it has a driver it doesn’t count... unless it’s Tesla. Then having a driver behind the system still counts!

I mean I literally just pointed out the opposite about Diplomats definition of "working"

But you keep blathering non-seq nonsense, it seems to be your thing.

 

[johnm]

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.

One problem with covering only certain cities is that people occasionally travel between cities. Having to transfer vehicles several times to make a trip seems like a not great experience.

When I see people talking about scaling, I find myself wondering how long it will take to cover substantially all of the 3,000 largest cities in the US [about 50% of the population] and thinking it won't be real soon. I guess if other people are thinking about scaling to 10 cities, that could probably be sooner.

 

[Bitdepth]

One problem with covering only certain cities is that people occasionally travel between cities. Having to transfer vehicles several times to make a trip seems like a not great experience.

That is all well and good but the average Uber ride is 5.41 miles, so that just depends on how far the cities are. A lot more people do not travel that far in an Uber or Lyft or Taxi in general. Moreover, people who use public transportation system, transfer buses and trains to get to their destination so i don't see why that wouldn't be acceptable for an autonomous taxi service. You have to think about the average mile of a trip.

When I see people talking about scaling, I find myself wondering how long it will take to cover substantially all of the 3,000 largest cities in the US [about 50% of the population] and thinking it won't be real soon. I guess if other people are thinking about scaling to 10 cities, that could probably be sooner.

I'm not concerned about a single company covering 3000 largest cities in the US, that is unfeasible at least not in the next several decades. No single company has to cover that amount alone. I'm more concerned with what they can realistically cover in the mid-term because that will prove the viability of the service. San Francisco is the 17th largest city in the US by population and Waymo and Cruise are proving their technology there by covering the entirety of it right now. The 10 largest cities have 20+ million population that's roughly 20% of the total Uber customers in 2021. I'm just looking at it from a more realistic POV.

 

[stopcrazypp]

That is all well and good but the average Uber ride is 5.41 miles, so that just depends on how far the cities are. A lot more people do not travel that far in an Uber or Lyft or Taxi in general. Moreover, people who use public transportation system, transfer buses and trains to get to their destination so i don't see why that wouldn't be acceptable for an autonomous taxi service. You have to think about the average mile of a trip.

Averages don't tell you much about the break down, as the shorter trips drag down the average. A distribution chart would be more telling.

For example, a very common trip via taxi or Uber is from SF to SFO airport. That is a 12 mile trip even from downtown, 16 miles if from the Richmond district.

I'm not concerned about a single company covering 3000 largest cities in the US, that is unfeasible at least not in the next several decades. No single company has to cover that amount alone.

That's entirely feasible if it's done with personal vehicles. It's only not feasible with fleets, and I disagree different companies would spread out to cover more area, rather they will all try to compete in the same most popular areas, just like how Waymo and Cruise are targeting the same areas (as did Uber and Lyft). The only way fleets would be spread out, with different companies covering different areas, is by sales territory agreements, which is illegal under federal law:
Market Division or Customer Allocation

Spreading well beyond the most popular cities is absolutely critical for people living in less populated areas, if they are to have any hope of experiencing self driving cars in their lifetime. That's why people keep bringing it up (and notice which states and cities they are from). Those of us living in the most popular metro areas probably have a different perspective than those people (but it's not hard to understand if you put yourself in their shoes).

I'm more concerned with what they can realistically cover in the mid-term because that will prove the viability of the service. San Francisco is the 17th largest city in the US by population and Waymo and Cruise are proving their technology there by covering the entirety of it right now. The 10 largest cities have 20+ million population that's roughly 20% of the total Uber customers in 2021. I'm just looking at it from a more realistic POV.

 

[diplomat33]

One problem with covering only certain cities is that people occasionally travel between cities. Having to transfer vehicles several times to make a trip seems like a not great experience.

highways is one of the domains Waymo is working on.

 

[diplomat33]

JJ Ricks takes a ride in Cruise! Not impressed IMO.

Cruise needed remote assistance to get around a pole in the road. Cruise also does some really jerky oversteering. Cruise oversteers on simple empty residential streets. Cruise has trouble lining up with certain medians and intersections. It actually reminds me of early FSD Beta. No wonder Cruise limits driverless to night time when traffic is simpler.

IMO, Cruise is way behind Waymo in how it drives.

Also, JJ suggested Cruise add a "FSD visualization" screen and gets shot down by the CEO. Ouch.

https://twitter.com/x/status/1603550801942781954

 

[diplomat33]

Here we see Cruise phantom brake hard in the middle of an empty road at night. Maybe the trailer parked on the side freaked it out? Really odd.

 

[powertoold]

IMO, Cruise is way behind Waymo in how it drives.

But they're both level 4 and driving in SF! Way better than Tesla's level 2!
/s

 

[Bladerskb]

But they're both level 4 and driving in SF! Way better than Tesla's level 2!
/s

They are... unlike you and Tesla fans, we can approach things critically and give out constructive criticism like JJRick did and not be hanged by an angry mob. It actually proves that real ADAS/AV lovers are NOT biased towards a particular company. They can logically deduce who is ahead aswell. IF another company shows up and deploys something better, we will logically assess and crown that company.

For Tesla fans like you, Tesla is 10 years ahead. No matter what.
Tesla FSD can plow through a crowd of children and kill dozens and still be considered 10 years ahead.
Nothing will alter your mind.
That's not logic or reason.
That's called fanaticism.

 

[stopcrazypp]

JJ Ricks takes a ride in Cruise! Not impressed IMO.

Cruise needed remote assistance to get around a pole in the road. Cruise also does some really jerky oversteering. Cruise oversteers on simple empty residential streets. Cruise has trouble lining up with certain medians and intersections. It actually reminds me of early FSD Beta. No wonder Cruise limits driverless to night time when traffic is simpler.

IMO, Cruise is way behind Waymo in how it drives.

Also, JJ suggested Cruise add a "FSD visualization" screen and gets shot down by the CEO. Ouch.

https://twitter.com/x/status/1603550801942781954

Is that how it is in general (including in SF) or is it from a result of overtuning to SF (in which case their bragging about only taking 90 days to launch in AZ is over exaggerated)? I don't recall similar complaints about SF, but then that may be because there haven't been the same type of review posted.

 

[diplomat33]

So @Bladerskb do you have any thoughts on the JJ video in Cruise? How do you explain the issues we see in the video?

 

[diplomat33]

Is that how it is in general (including in SF) or is it from a result of overtuning to SF (in which case their bragging about only taking 90 days to launch in AZ is over exaggerated)? I don't recall similar complaints about SF, but then that may be because there haven't been the same type of review posted.

There are a couple long unedited Cruise videos in SF that you can find on reddit but for the most part, almost all the videos we see from Cruise in SF are highly edited. They are usually just 1 mn "tick tock" style videos of someone going "OMG, the car has no driver! I am living in the future". And of course, Cruise does a great PR job of retweeting these videos to promote that they are deploying true driverless cars. However, JJ always does a great job with these reviews because, like with his Waymo videos, he will share longer unedited videos that show the good and the bad. He does much more raw and comprehensive reviews of the tech, wheter it is Waymo or Cruise.

It could be that Cruise actually is that bad in SF and they just do a good PR job of hiding it. And it could be that Cruise limits driverless at night precisely to make the ODD easier because the tech is not ready for more complicated day driving. This JJ video and the reports of Cruise stalling and blocking traffic seem to support that hypothesis.

It could also be that Cruise is super overfit to SF and so the tech works better in SF than in other places. The fact that they have spent 9 years developing autonomous driving only in SF and only recently expanded their autonomous driving outside of SF would seem to support that hypothesis.

It could also be that Cruise rushed deployment to Phoenix and Austin because they wanted to keep their timeline of deploying in 90 days. Cruise could be under a lot of pressure from GM to show that they can scale. All that pressure might have led Cruise to cut corners and maybe not test as much as they should or maybe not map certain areas. This could also explain the poor performance.

Personally, I am really surprised and kinda shocked. Driving empty residential streets is very easy. I expectected better from Cruise which has been working on AVs for 9 years.