DanCar
Active Member
Lots of electric Waymo Jaguar I-Paces running around in SF.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Waymo Says Its Self-Driving Tech Is More Advanced Than Tesla FSD, Others
- Thread starter TMC Staff
- Start date
diplomat33
Average guy who loves autonomous vehicles
Waymo has the Jaguar I-Pace which is all electric.
Yes, they can undercut Uber. See the pic below from a Waymo ride. We can see the cost was $12.41 for 9.3 mile trip. That's $1.33/mi. The average cost of an Uber is $2/mi. So Waymo's cost per mile is already less than the average cost per mile of an Uber.
I'm sure the guy bringing his Waymo car will brag about how his "more advanced system" costs Wayless than the $10k Tesla is charging for FSD. 
HighZ
Member
I doubt that the hardware cost is being amortized into that per mile price. It would seem that it is being subsidized by Waymo initially.
Waymo's business is a self-driving taxi service, so indeed you can not purchase a car with Waymo's self-driving system. They have, however put it out already.
You can currently ride in a fully-self-driving L4 Waymo in a geofenced area; you can not find fully-self-driving Teslas in any area. Whether you can purchase a car or otherwise is an irrelevant reach for the sake of being in Tesla's defense.
I do however agree with the sentiment behind "apples to oranges". While Waymo does provide full-self-driving, it is currently severely limited. Tesla's entirely different approach hasn't hit this milestone yet, but it could reach greater generality with acceptable performance sooner than Waymo.
Don't most Uber rides take place in densely populated areas. I think the geofencing thing is not even an issue. Shouldn't there always be some geofencing? I don't want my fsd tesla driving from NYC to Florida on me.
Waymo could geofence all of Manhattan it's only 22 miles. I don't think service in the suburbs would be that profitable anyway.
Waymo could partner with NYC and drive buses. They drive the same continuous route over and over. Sometimes in bus lanes. There is a bus that goes 1.8miles and makes three stops before circling back. B39
Waymo could geofence all of Manhattan it's only 22 miles. I don't think service in the suburbs would be that profitable anyway.
Waymo could partner with NYC and drive buses. They drive the same continuous route over and over. Sometimes in bus lanes. There is a bus that goes 1.8miles and makes three stops before circling back. B39
diplomat33
Average guy who loves autonomous vehicles
It's not just ride-hailing. According to Krafcik, Americans do most of their daily driving in a 100 sq mi area. Incidentally, that might be why Waymo picked 100 sq mi for their geofenced area in Chandler. And, according to driving stats I found, the average work commute is 16 miles. So I think a case could be made that geofencing could still be very useful to a lot of Americans, even for personal robotaxi ownership, not just for ride-hailing. And if you expand the geofencing to include all major metro areas in the US and major highways, it would allow L4 to do most long distance travel as well. Yes, it would still exclude people living in rural areas but like you said, most demand in in urban areas. For companies like Waymo looking to sell driverless rides and eventually lease robotaxis to consumers, there will be plenty of profit in dense urban areas.
In fact, I think it could be argued that L5 is not really needed, at least not right away. Depending on where and how big the geofenced area is, L4 could still be good enough for most Americans' driving habits. The bigger the geofenced area gets, the more useful it gets. Eventually, your geofenced area gets big enough and close enough to L5 ODD that it is "good enough" for most people.
L4 is fine for lots of people as long as it's L4 in all weathers at a price low enough to eliminate a car. It can't just be a cheap taxi. It needs to be a better bus.
I'm in the camp of Apples to Oranges. I've been following Waymo (Google ) since 2012. I've also been following Tesla/Elon since 2006. I've seen everyone ridicule him and call multiple things impossible and then moving the goal post. (I'm sure there are 10,000 thread on here about this.)
With that being said, Living in a small city in Ohio, I have driven 22,500 miles on my M3 in 8 months since I got it in March. I would say that I have had autopilot engaged for at least 19,000 of those miles. I wouldn't be able to do that with SuperCruise or any other L2 ADAS
1. I cannot get a waymo ride (or even an Uber) in my town. But looking at the FSD beta videos, I can see that driving me to work every day with no problems. Hopefully I will find out in another couple of months.
2. I really love watching the Waymo videos because it really is the first level 4 robotaxi available for consumers and they should be applauded for that achievement, but that really doesn't mean anything to me if I can't use it at all and probably won't be able to use it for a very long time.
3. Waymo has no obligation to prove anything to us, BUT if they are trying to throw shade Tesla's way, and they really are superior, I would like to see them take a car somewhere that isn't pre-mapped and let it drive from one point to another. UNTIL that happens, must say that I am more impressed with what I've seen from Tesla FSD than Waymo.
With that being said, Living in a small city in Ohio, I have driven 22,500 miles on my M3 in 8 months since I got it in March. I would say that I have had autopilot engaged for at least 19,000 of those miles. I wouldn't be able to do that with SuperCruise or any other L2 ADAS
1. I cannot get a waymo ride (or even an Uber) in my town. But looking at the FSD beta videos, I can see that driving me to work every day with no problems. Hopefully I will find out in another couple of months.
2. I really love watching the Waymo videos because it really is the first level 4 robotaxi available for consumers and they should be applauded for that achievement, but that really doesn't mean anything to me if I can't use it at all and probably won't be able to use it for a very long time.
3. Waymo has no obligation to prove anything to us, BUT if they are trying to throw shade Tesla's way, and they really are superior, I would like to see them take a car somewhere that isn't pre-mapped and let it drive from one point to another. UNTIL that happens, must say that I am more impressed with what I've seen from Tesla FSD than Waymo.
If one good thing comes from this pandemic let's hope that it's an end to the daily commute for millions of people.
Waymo are of course going to be cheaper than Uber. Uber has to pay for human drivers and robotaxis can run 24/7 in principal, charging aside.
JSML
Member
Waymo hasn't solved all the tail problems even within their geo-fenced area. Aborted rides and remote driving still happens. My question is will they be able to solve all the problems or will it always be supported and remotely supervised? Waymo also doesn't operate on highways yet. Will they be confident enough to put their taxis on the highway and assume the liabilities?
Those are great questions, and even if it can't handle every situation, if its able to alert remote drivers when needed, you could have 1 driver per 100 vehicles that are deployed let's say and drive down costs for robotaxis
diplomat33
Average guy who loves autonomous vehicles
Sure, they will solve all them in time. I don't see why not. The fact that they are offering driverless rides implies that they have achieved several 9's of reliability already, probably to 99.999%. So I don't see why they can't solve the last 9's too. Waymo has the top Google engineers, the best hardware, the best machine learning.
Waymo did operate on highways before but temporarily stopped. I am sure they will be back on highways soon. Waymo robotaxis are able to handle highway driving.
JSML
Member
diplomat33
Average guy who loves autonomous vehicles
Not really. I would be making an educated guess, at best.
But I would point out that you actually don't need to solve ALL problems. You only need to get FSD to 99.9999% in order to be good enough for deployment without driver supervision.
But I doubt we will even agree on what constitutes "99.9999%" so it is a moot point to try to set timelines.
We do know that Waymo has fully driverless rides now in a 50 sq. mi area. This implies that Waymo has achieved "99.9999%" in the 50 sq mi. We know their next most immediate goal is to offer fully driverless rides in a 100 sq mi area. IMO, when they achieve that, I would consider Waymo to have achieved "99.9999%" in the 100 sq mi geofenced area. Then, their next goal will be to launch similar driverless geofenced ride-hailing in other metro areas like LA or SF. I have no idea what the time frame for that might be.
diplomat33
Average guy who loves autonomous vehicles
By the way, we don't need to guess at what long tail problems Waymo still needs to solve. We can look at the Waymo safety data to get a sample of the long tail problems. We see that in 6M miles, Waymo did not have any perception problems. It did not have any issues with traffic lights, stop signs or navigating intersections or roundabouts. It never missed a turn. It never had an accident where it left the lane. All the problems involved reacting to other vehicles, and most of them were caused by the other driver behaving badly.
For example, there were cases where another car was speeding and rear ended a Waymo. There were cases where a car drove in the wrong direction and risked a head on collision with a Waymo. There were a couple cases of Waymo almost side swiping a car that moved too quickly into a left lane when the Waymo was trying to merge left. There was a case of a car that ran a red light and would have hit the Waymo going through an intersection. There was another case of a car failing to yield to a Waymo and would have hit the Waymo. There was a case of a truck that failed to stop at a stop sign and would have hit the Waymo. There was another case of a car that was occluded, that made a turn in front of the Waymo.
Now, some of these cases might have been unavoidable. We know it is not always possible to avoid all accidents. But I believe these are cases that Waymo can "solve". Waymo can reduce the risk of these situations becoming a real accident by improving the driving policy and improving defensive driving. And the good news is that these cases that are not unique to this geofenced area. They happen in driving all over the US. So once Waymo does improve the car's driving to handle these cases better, it will improve Waymo in general, in all areas where Waymo operates.
EVNow
Well-Known Member
But ***only in the places they are testing ***. A very big caveat. This is the problem with Waymo's approach.
diplomat33
Average guy who loves autonomous vehicles
Waymo is testing in 25 metro areas across 11 States in the US. So it is a big and diverse testing area. Waymo is not just testing in a small geofenced area. The commercial robotaxi service in Chandler is NOT the entire testing area that Waymo is using.
mikes_fsd
Banned
Tesla FSD is publically testing in 13 states (??plus DC??) without ANY geo-fencing to the testers
Tesla must be the new leader in all of autonomy because not only do they test, they let all of their testing be broadcast to the entire world (the good, bad, ugly, and plain ludicrous).
I wish these pretend autonomy co's like Waymo would be so transparent with their testing. I mean c'mon!
https://twitter.com/28delayslater/status/1329624089292660737
diplomat33
Average guy who loves autonomous vehicles
How many cars have FSD Beta in these 13 States?
I wish these pretend autonomy co's like Waymo would be so transparent with their testing. I mean c'mon!
Waymo has published the most transparent, most detailed safety report of any autonomous company of their L4 autonomy showing how their FSD works, how they test, and every incident, every case, good and bad. The report covers 6M autonomous miles which is more data than any other autonomous company has published so far. But Tesla is more transparent because they are letting a couple fans on youtube publish videos???!?!? Mmmm...okay... LOL.
Similar threads
