Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Waymo One doubles service area in Phoenix, continues growing in SF

blog.waymo.com

Post reply on HN
Print view View original

511–520 of 520 posts

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#511

Post by fastball »

ra7 wrote: post #507
fastball wrote: post #506

Earlier quoted context omitted.

I'm not equating it with solving the entire problem, I am pushing back against the idea that autonomous driving can be sub-divided into completely discrete sub-problems. So I guess you could say no, I do not understand that sub-problems can be solved. Object detection is inextricably linked with prediction which are both inextricably linked with planning.

Of course, they are linked. It doesn't mean you need to operate worldwide to prove sensor fusion is working. There is no scaling issue with it. Sensor fusion is already proven effective as companies using it are showing excellent safety records where they are operate. The Tesla complaints about "what do you do in case of disagreements?" don't hold good anymore. They just don't want to invest in get it working (but ar…

What evidence do you have that there is no scaling problem with sensor fusion?

As you point out, the companies that have "solved" it don't have scale.

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#512

Post by ra7 »

fastball wrote: post #511
ra7 wrote: post #507

Earlier quoted context omitted.

Of course, they are linked. It doesn't mean you need to operate worldwide to prove sensor fusion is working. There is no scaling issue with it. Sensor fusion is already proven effective as companies using it are showing excellent safety records where they are operate. The Tesla complaints about "what do you do in case of disagreements?" don't hold good anymore. They just don't want to invest in get it working (but ar…

What evidence do you have that there is no scaling problem with sensor fusion? As you point out, the companies that have "solved" it don't have scale.

Sensor fusion has nothing to do with scale. They are merely algorithms that combine multiple real time inputs into a single low level representation. This process remains the same whether you operate in San Francisco or worldwide. How they are interpreted is what differs from a Waymo to a Cruise to a Tesla. Waymo/Cruise get richer input they can feed into their downstream perception algorithms. Look at Waymo’s object detection benchmarks against Argoverse, KITTI and Waymo Open Dataset for their performance. They top most of them.

They don’t have scale yet for different reasons (adverse weather driving, long testing cycles, costly operations setup). “If they don’t have scale, that must mean sensor fusion doesn’t work” is quite a ridiculous argument.

If you think sensor fusion is not possible, then you must also believe Tesla’s attempts at re-integrating radar will not work and that they won’t achieve their stated goal of FSD.

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#513

Post by NewMountain »

asvitkine wrote: post #311
NewMountain wrote: post #111

Earlier quoted context omitted.

I had an Uber driver in upstate New York who was trying to get me to agree or at least not disagree with the superiority of the white "race". I had another driver in Oakland California who would not relent until I acknowledged God is real and I should be thankful for his love. I felt massive cognitive dissonance afterwards feeling both the requirement to report these drivers but also massive guilt at the non-zero cha…

Wouldn't such cases result in bad ratings for the drivers?

I honestly don't recall ever having a driver lower than 4.2 or so ever when taking an Uber. My guess is both of those drivers were low to mid 4 stars as I didn't pay any attention to it.

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#514

Post by StrictDabbler »

bmelton wrote: post #260
StrictDabbler wrote: post #171

Earlier quoted context omitted.

I have never tried to spell the phrase "flip a youey" before. It might be the only word I use that I have never considered spelling, so I am grateful that you included it your comment. That said, I think you are wrong and it should be "uey", "uie," "u-ie," or any other spelling that emphasizes the "u" after which the term is named. Also "I do not know the answer but I am sure you are wrong" is the most internetish co…

> I have never tried to spell the phrase "flip a youey" before. As a former Scripps spelling bee contestant, I've developed a low-grade allergy to commonplace onomotopeic words with no obvious, agreed-upon spelling. Recently I struggled with zhuzh (which has the bonus terror of having an alternate spelling, 'zhoosh') - but I commend OP for having chosen a phonetically parseable spelling over one of your (arguably mor…

I had always imagined that word to be spelled "jouerge", a variation on the French verb "jouer: to play" mixed with some element of "bourgeois: in the manner of the middle class".

My etymology makes sense. You can even extend it to the "jouergeois" if you need to describe middle-class foodies.

In this case I am definitely wrong and I blame Canada.

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#515

Post by sib »

agensaequivocum wrote: post #209
sib wrote: post #163

Earlier quoted context omitted.

Phoenix is like LA without the positives

This is definitely not true.

Like LA: Large, sprawling, full of concrete, car-dependent Without the good parts: No media industry, very little tech industry, much less diverse, far fewer good restaurants, no ocean

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#516

Post by fastball »

ra7 wrote: post #512
fastball wrote: post #511

Earlier quoted context omitted.

What evidence do you have that there is no scaling problem with sensor fusion? As you point out, the companies that have "solved" it don't have scale.

Sensor fusion has nothing to do with scale. They are merely algorithms that combine multiple real time inputs into a single low level representation. This process remains the same whether you operate in San Francisco or worldwide. How they are interpreted is what differs from a Waymo to a Cruise to a Tesla. Waymo/Cruise get richer input they can feed into their downstream perception algorithms. Look at Waymo’s object…

I never said I don't think sensor fusion is possible. My only argument has always been "adding more sensors does not mean you will have a better self-driving car". Richer data does not mean better data. It does not mean more actionable data either. All it means is that you have more data points. I absolutely agree that this can be a good thing, and probably is in the case of self-driving.

But maybe it's not. Maybe the additional flops required for sensor fusion would be more useful if given over to the planning part of the problem. After all, you don't have infinite cycles. This is what I mean when I say it has not been proven at scale. The edge cases matter (arguably they are all that really matter), and "solutions with more sensors work better at object detection in artificial benchmarks" is not nearly as convincing to me as it seems to be to you. Again, object detection does not exist in a vacuum, and a solution that generally behaves better in a very constrained environment might actually behave worse when you throw all the edge cases at it and add on prediction and planning.

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#517

Post by ra7 »

fastball wrote: post #516
ra7 wrote: post #512

Earlier quoted context omitted.

Sensor fusion has nothing to do with scale. They are merely algorithms that combine multiple real time inputs into a single low level representation. This process remains the same whether you operate in San Francisco or worldwide. How they are interpreted is what differs from a Waymo to a Cruise to a Tesla. Waymo/Cruise get richer input they can feed into their downstream perception algorithms. Look at Waymo’s object…

I never said I don't think sensor fusion is possible. My only argument has always been "adding more sensors does not mean you will have a better self-driving car". Richer data does not mean better data. It does not mean more actionable data either. All it means is that you have more data points. I absolutely agree that this can be a good thing, and probably is in the case of self-driving. But maybe it's not. Maybe th…

> Richer data does not mean better data. It does not mean more actionable data either.

How actionable you make it depends on how good your engineers are. Diverse input data is definitely more actionable than just RGB pixels from images.

> But maybe it's not. Maybe the additional flops required for sensor fusion would be more useful if given over to the planning part of the problem. After all, you don't have infinite cycles.

On the contrary, it reduces your compute requirements. For example, adding lidar means you are getting direct distance measurements instead of wasting compute cycles predicting distance using voxels. It's a much more effective use of compute.

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#518

Post by fastball »

ra7 wrote: post #517
fastball wrote: post #516

Earlier quoted context omitted.

I never said I don't think sensor fusion is possible. My only argument has always been "adding more sensors does not mean you will have a better self-driving car". Richer data does not mean better data. It does not mean more actionable data either. All it means is that you have more data points. I absolutely agree that this can be a good thing, and probably is in the case of self-driving. But maybe it's not. Maybe th…

> Richer data does not mean better data. It does not mean more actionable data either. How actionable you make it depends on how good your engineers are. Diverse input data is definitely more actionable than just RGB pixels from images. > But maybe it's not. Maybe the additional flops required for sensor fusion would be more useful if given over to the planning part of the problem. After all, you don't have infinite…

You are getting direct distance measurements for some things, but not for everything, so you still need a system that is predicting distance using voxels.

LIDAR doesn't work great in the rain/snow, so you still need a system that is predicting distance using voxels.

Doesn't really sound like less compute to me. I feel like I'm repeating myself but again this is what I mean by scale matters. A self-driving system that performs better in perfect conditions is not even a good system if it doesn't work in the rain.

Which is the point the detractors of "fuse all the things" have been making. Don't get me wrong, I think probably those same detractors (e.g. Elon Musk) have also not been entirely intellectually honest with their arguments and have ulterior motives like cost savings and saving face, but that doesn't make the arguments wrong.

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#519

Post by sprout20 »

addisonl wrote: post #504
tbruckner wrote: post #135

Earlier quoted context omitted.

How is pricing?

It’s definitely not free, I found rides to be as expensive or more expensive than an Uber/Lyft.

Jumping back to this thread. My understanding is that the California Public Utilities Commission is the body that regulates this and Waymo does have the approval to provide driverless rides at no charge (Drivered Deployment) but does not yet have the approval it needs to charge for those rides (Driverless Deployment). I think Cruise is the only service that's allowed to charge for fully driverless rides in SF so far.

https://www.cpuc.ca.gov/news-and-updates/all-news/cpuc-appro...

https://www.cpuc.ca.gov/regulatory-services/licensing/transp...

Re: Waymo One doubles service area in Phoenix, continues growing in SF

#520

Post by sprout20 »

mustacheemperor wrote: post #129
sprout20 wrote: post #93

I've had access in SF for the past few months and have been really impressed! The Waymo rides I've taken have been confident + a bit conservative (e.g. driving at the speed limit when traffic is flowing more quickly), but overall the team's done a great job with the experience. It's free for now so I've been using it regularly during commuting/rush hour and casual drives. Some quick thoughts: Advantages over Uber/Lyf…

Did you get access from the public beta sign-up list, or some other channel? I've been on the waitlist for over a year and have seen people utilizing Waymo near my apartment, but nothing's budged for me. I did get into the Cruise beta, and that just made me positive Waymo will almost completely replace other forms of car transit for me once it's available, but Cruise is very limited.

Just public beta sign-up list as far as I know! I don't remember doing anything special besides signing up for the waitlist and filling out a survey, so I guess I just got lucky.
Post reply on HN