Waymo is opening its driverless service to more people in Phoenix • Hacker News Board

Re: Waymo is opening its driverless service to more people in Phoenix

#201

Post by cloudwalking » Fri, Oct 09, 2020, 2:41 PM UTC

jimmydorry wrote: post #167

tobyjsullivan wrote: post #110

Earlier quoted context omitted.

I imagine you're correct that lidar is a huge leg-up in the race to self-driving. However, I can see a scenario where, despite that, it is not the right choice for Tesla. It's important to remember that self-driving technology will be commodity at some point (like the lidar units themselves). Whoever offers the most affordable option meeting some baseline performance will dominate the market and licence to every self…

I can't imagine waymo or any of the other groups that are basing their technology around lidar, managing to get to "commodity stage". LIDAR has a fixed range it is useful for, which is greatly shrunk when you also need to factor in vehicle response time. Cameras can see as far as humans can (further actually, as it can see things more reliably at greater distances than humans). Going beyond 25mph is already massively…

This is simply not true, as evidenced by Waymo driving on roads faster than 25 mph

Re: Waymo is opening its driverless service to more people in Phoenix

#202

Post by Ajedi32 » Fri, Oct 09, 2020, 4:15 PM UTC

ma2rten wrote: post #181

flokie wrote: post #28

I know the community is generally very negative/skeptic on this (at least based on the 21 comments so far below), but this is an amazing technical & business accomplishment. Congratulations! The blog post is relatively clear that ramp up will take a few weeks - "we’ll start with those who are already a part of Waymo One and, over the next several weeks, welcome more people directly into the service through our app" b…

* this is an amazing technical & business accomplishment* No, it's not. Waymo has been doing amazing work, but the problem is that they PR department is a little bit too good at it's job. Waymo has these kinds of announcements very couple of months (sometimes more often) and if you don't pay close attention all of them sound like Waymo finally had a breakthrough. But if you look closer each of them is only small incr…

Yes it is. True, this isn't some sudden breakthrough; just the culmination of many years of slow and steady progress. That doesn't make it any less of an achievement though. This is the first time ever that fully driverless cars are going to be available to members of the general public who aren't under an NDA.

Waymo is taking a slow but steady approach to the development and roll-out of self-driving cars, and I agree with you that that can sometimes make the individual milestones they choose to announce along the way seem pretty minor. Milestones are arbitrary by nature after all, and Waymo has a long road ahead before driverless cars become ubiquitous. Over time though, this sort of slow, incremental progress is going to add up.

I made this same point a couple years ago, when Waymo first announced Waymo One[1]. The point where "self driving cars are here" is going to be a blurry line.

[1]: https://news.ycombinator.com/item?id=18609982

Re: Waymo is opening its driverless service to more people in Phoenix

#203

Post by Ajedi32 » Fri, Oct 09, 2020, 4:23 PM UTC

polishTar wrote: post #111

sandworm101 wrote: post #46

Earlier quoted context omitted.

"Members of the public service" is different than "members of the public". This is still a closed beta. But now it seems you are allowed to have your friends along for the ride. I wonder if these friends and family persons are subject to any NDA.

Waymo One doesn't use NDAs and users are allowed to bring friends with them, but Waymo also has an "early rider" program that's more restrictive and does use NDAs.

Correct. This article is about "fully driverless" rides (rides with no safety driver present) graduating from their experimental "Early Rider Program" (which includes a small group of people under NDAs) to their public "Waymo One" service (which is a commercial service which requires no NDAs, but still has rather limited availability for now).

Waymo One has been shut down for a while now due to COVID, but now that driverless rides are an option they're re-opening the service with 100% fully driverless rides. Once they get around to installing barriers between the driver and passenger seats they plan to go back to most rides having a safety driver, but with some smaller percentage of Waymo One rides continuing to be fully driverless.

Re: Waymo is opening its driverless service to more people in Phoenix

#204

Post by gbronner » Fri, Oct 09, 2020, 4:25 PM UTC

kanwisher wrote: post #194

gbronner wrote: post #160

Earlier quoted context omitted.

Hardware costs will go way down. Lidar @ 100$, cheaper chips, cheaper sensors, etc. They are still in a learning phase -- they charge enough so as not to distort rider behavior, but they expect to lose money on each ride. Apart from that, why build cars when you can sell the system to everyone?

Lidar requires accurate maps, using vision allows Tesla to drive on roads its never seen

You can use both if you want to. Tesla went with cameras only because they thought Lidar was too expensive.

Not having lidar has directly caused a bunch of autopilot crashes, so I expect they'll add it.

Re: Waymo is opening its driverless service to more people in Phoenix

#205

Post by chubot » Fri, Oct 09, 2020, 4:53 PM UTC

jryle70 wrote: post #140

chubot wrote: post #120

Earlier quoted context omitted.

Weird question :) My claim is that it won't be open to "anyone" in a useful way any time soon. And you need to define "anyone" -- that's precisely what's being fudged here (and what's been fudged in previous announcements). To be clear: I think Waymo will fold or be wound down before this happens. In a previous comment, I said that IF we get economically feasible self-driving cars in the next 10-20 years, then it wil…

> Weird question :) My claim is that it won't be open to "anyone" in a useful way any time soon. Seems pretty clear it will be available to public in the next few weeks, unless you're saying "general access to anyone who chooses to download the app" isn't about general public. "Beginning Thursday, any existing Waymo One customer can hail a driverless minivan from a fleet of more than 300. The vehicles will be operati…

Again, "anyone" needs to be defined.

Headline: "Waymo is opening its fully driverless service to the general public in Phoenix"

But

Beginning Thursday, any existing Waymo One customer can hail a driverless minivan from a fleet of more than 300. The vehicles will be operating in a smaller, roughly 50-square-mile service area.

Qualifications:

(1) existing Waymo One customers

(2) you can't go anywhere in Phoenix. Phoenix is a lot bigger than 50 square miles. (Also, even if the area were bigger, 300 cars sounds like a small number to cover it)

(3) Waymo plans to reintroduce safety drivers for some rides as it expands its Phoenix service area

Is it driverless or not?

I'm trying to read between all the doublespeak, and I don't see much progress. Again, I would love for someone who lives in Phoenix to sign up, try it, post here, and prove me wrong. Tell me what it was like.

Re: Waymo is opening its driverless service to more people in Phoenix

#206

Post by whimsicalism » Fri, Oct 09, 2020, 5:07 PM UTC

disgruntledphd2 wrote: post #192

whimsicalism wrote: post #48

Earlier quoted context omitted.

I agree that it is a very different model, but it's a bit much to say that externalizing car repair costs is what the business model of Uber/Lyft is "based on"

Dunno if you'll see this, but when I made my comment, this link was much of my source: https://ftalphaville.ft.com/2019/04/29/1556557142000/Scoop--...

Even that paper demonstrates that if you weaken their assumptions about taxi-car occupancy being exactly the same as Uber/Lyft and don't assume that you will have a remote monitor per every 10 cars (unsure why they would include this assumption, even the current Waymo setup doesn't have this), self-driving actually does come out as priced less.

They also don't source their cost for CDV, which is substantially lower than any other number I can find cited in paper. Further, the current regulatory framework includes huge wealth transfers/incentives for car ownership, so it's not a fair comparison. For instance, the $130/yr cost of a parking permit in SF is a huge wealth transfer to car owners as that is substantially below what the market-price would be for that permit.

Re: Waymo is opening its driverless service to more people in Phoenix

#207

Post by klmadfejno » Fri, Oct 09, 2020, 8:49 PM UTC

kevincox wrote: post #124

klmadfejno wrote: post #96

Earlier quoted context omitted.

* You don't need to pay drivers * You don't need an army of marketers/hr/legal/support staff for your drivers * You don't need to dodge ambiguous contractor vs. employee laws * You don't have any risk about individual drivers being bad actors or acting fraudulently * You can predict, with great accuracy, how many cars will be available at any given time, whether or not they'll want to do another drive after (they wil…

There is also a direct benefit from riders who don't want to interact with a driver. Or want to have conversations that they would be uncomfortable subjecting a driver to. I mean sure, there is a camera. But that is hugely different than a person (both positive and negative)

I wonder if that'll play poorly in the business crowd. Talking in hushed tones is bad but passable when there's a human driver. It's terrible when it's a recording device.

Re: Waymo is opening its driverless service to more people in Phoenix

#208

Post by kevincox » Fri, Oct 09, 2020, 8:57 PM UTC

klmadfejno wrote: post #207

kevincox wrote: post #124

Earlier quoted context omitted.

There is also a direct benefit from riders who don't want to interact with a driver. Or want to have conversations that they would be uncomfortable subjecting a driver to. I mean sure, there is a camera. But that is hugely different than a person (both positive and negative)

I wonder if that'll play poorly in the business crowd. Talking in hushed tones is bad but passable when there's a human driver. It's terrible when it's a recording device.

It's funny. For fun personal things I don't want to annoy the driver, but don't really care if what I say is recorded. For business you don't really care if the driver hears it, but don't want to get recorded.

I suspect that a well-defined retention and usage policy could satisfy most business. In fact they might prefer it to having a driver listening in.

Re: Waymo is opening its driverless service to more people in Phoenix

#209

Post by petra » Sat, Oct 10, 2020, 9:15 PM UTC

lhorie wrote: post #152

petra wrote: post #123

Earlier quoted context omitted.

Yes. There are other important factors. But at least for the next few years, technology is the barrier. So Google can partner with Lyft, for example, to get a good brand, app, customer service, regulation, etc. But on top of that, the brand in this market is the promise "our cars are safe". Google is in a distinct place to own that, because of all the press they had, they press they will have, and by really leading t…

I'm not actually convinced that technology is necessarily a barrier. People take ubers today. Heck, people take taxis today. I think a bigger challenge for Waymo is going to be getting trust and regulatory favor. Everyone keeps clinging on to the "oh, comparing prices, waymo is cheaper than uber because there's no driver!" narrative, but from what I've seen from the few reports so far, the prices aren't all that diff…

>> but from what I've seen from the few reports so far, the prices aren't all that different from other services.

Current prices include a safety driver and small scale. Prices should go down in the future.

And sure, regulation and trust should be huge barriers. But besides the customer service issue(which seems too pessimistic), Google should be leading over others, on the trust issue(no accidents, more miles, highly skilled tech company), And on the regulation issue(good ties with Washington).

But those 2 are still a big risk, like you mention.

Re: Waymo is opening its driverless service to more people in Phoenix

#210

Post by donor20 » Sun, Oct 11, 2020, 10:06 AM UTC

tobyjsullivan wrote: post #110

donor20 wrote: post #102

Fantastic. The reality is that good mapping and some level of lidar (which is getting cheaper faster for low res/angle sensors) is somewhat game changing in terms of self driving vehicle reliability. Tesla will miss out I think by not taking advantage of doing even the cheapest possible one sensor lidar to augment their existing system. $500 -> in the future $100 for the data lidar gives you is gold.

I imagine you're correct that lidar is a huge leg-up in the race to self-driving. However, I can see a scenario where, despite that, it is not the right choice for Tesla. It's important to remember that self-driving technology will be commodity at some point (like the lidar units themselves). Whoever offers the most affordable option meeting some baseline performance will dominate the market and licence to every self…

This is not true. Google is collecting incredible levels of ground / lidar truthed imagery. In terms of then doing a generation function from imagery to 3D model (without lidar) they are going to be in a good spot if they want to be. This is the same approach they have taken in many many other areas - huge ground truthed datasets, then infer without the ground truth.