Waymo Starts To Open Driverless Ride-Hailing Service To the Public (techcrunch.com) 41
Waymo, the Google self-driving-project-turned-Alphabet unit, is beginning to open up its driverless ride-hailing service to the public. From a report: The company said that starting today members of its Waymo One service will be able to take family and friends along on their fully driverless rides in the Phoenix area. Existing Waymo One members will have the first access to the driverless rides -- terminology that means no human behind the wheel. However, the company said that in the next several weeks more people will be welcomed directly into the service through its app, which is available on Google Play and the App Store. Waymo said that 100% of its rides will be fully driverless -- which it has deemed its "rider only" mode. That 100% claim requires a bit of unpacking. The public shouldn't expect hundreds of Waymo-branded Chrysler Pacifica minivans -- no human behind the wheel -- to suddenly inundate the entire 600-plus square miles of the greater Phoenix area. Waymo has abut 600 vehicles in its fleet. About 300 to 400 of those are in the Phoenix area. Waymo wouldn't share exact numbers of how many of these vehicles would be dedicated to driverless rides. However, Waymo CEO John Krafcik explained to TechCrunch in a recent interview, that there will be various modes operating in the Phoenix area. Some of these will be "rider only," while other vehicles will still have train safety operators behind the wheel. Some of the fleet will also be used for testing.
bad time to be an Uber driver (Score:1)
Re: (Score:2)
Re:bad time to be an Uber driver (Score:4, Interesting)
Yeah it's amusing that as of like 2 days ago people were still saying that self driving cars were impossible and would never work in Slashdot comments.
This is great news for Tesla. It proves that it's theoretically possible to make Autopilot work. Although the Tesla approach and the Waymo approach are so different that it doesn't yet confirm that Tesla's big data, sensor/map light approach specifically will pan out.
Usually though the largest impediment to widespread adoption is a single proof of concept working.
Re:bad time to be an Uber driver (Score:4, Insightful)
The different strategies are fascinating to watch.
Toyota and Mercedes are smart and started early but have proceeded so cautiously, they'd never get there. They're so big they have too much to lose.
Uber went at it aggressively and ineptly. Fail.
Tesla has proceeded remarkably far by being aggressive and using their customers as beta testers for progressive automation. They've made it work and might win.
Waymo started early and has poured money into research and testing, and has chosen to focus on full automation rather than assists. They seem on track to get there first, but there is no partial success for them. If the research stalls out before full autonomy is viable, they lose.
Re:bad time to be an Uber driver (Score:4, Interesting)
Waymo and everyone else anywhere near to full self driving uses lidar. Volvo and others are fitting lidar to cars next year.
Tesla doesn't have lidar.
Re: (Score:2)
"Toyota and Mercedes are smart and started early but have proceeded so cautiously,"
You know how it is, things to do, people to defraud ...
Re: (Score:2)
Now
Re: (Score:2)
Can it handle 100% of all driving situations without human assistance 100% of the time in a way that is at least 100% as competent as a competent human driver? If 'no' then it's a piece of shit that shouldn't be allowed on public roads or be responsible for human lives.
And I mean ALL DRIVING SITUATIONS. No exceptions.
It can't because IT CANNOT 'THINK'. Has no capacity to 'think', and never will, until we understand what 'thinking' actually is, and we're nowhere near that yet.
I won't accept any argu
Re: (Score:2)
If 'no' then it's a piece of shit that shouldn't be allowed on public roads or be responsible for human lives.
I won't accept any arguments about it being 'less lethal than human drivers'
Why? Aren't you somewhat contradicting yourself? It is a reasonable criterion IMO.
Anyway, I pretty much agree with you about how scary it is to put so much control in the "hands" of AI, at least for now. Let's consider lethality. Even if AI turns out to perform statistically better than human drivers overall, I think it will be quite hard for society to accept any failure modes that would not apply to a human driver. For instance a sensor could be blinded by the sun and fail to detect a pedestrian, and th
Re: (Score:2)
How many screw-ups or deaths does a particular model of SDC have to have before it's banned? 2? 3? Infinite?
Re: (Score:2)
Re: (Score:2)
Good time to be an Uber driver. You'll soon be doing something more useful with your life instead of being essentially a scab taxi driver without the pay and benefits.
Every advance of technology that has eliminated menial jobs in the past several hundred years has turned out to be a good thing for those workers in the long run. There's no reason to think that track record suddenly changes now.
and will riders have to face criminal consequences (Score:2, Interesting)
and will riders have to face criminal consequences for an crashes like the uber safety driver?
Be on the hook for an DUI if in driver less mode (as stop button may count as in control under the DUI law)
Re: (Score:2)
No.
Re: (Score:1)
Until someone dies (Score:2)
Meanwhile the tens of thousands of deaths per year by human drivers don't get the same reaction.
Re: (Score:2)
I'm sure Google has thought through this PR scenario.
Re: (Score:2)
They probably waited until the 100% driverless cars passed 113 million miles (the current fatality rate) before this rollout. So now if anyone dies they can immediately say "Even with this tragic death we are still safer than a human driver."
Re: (Score:2)
Apples Oranges (Score:2)
Re: (Score:2)
Re: (Score:2)
Self-driving cars have already killed people.
The sky didn't fall. People are not as stupid as you think they are.
List of self-driving car fatalities [wikipedia.org]
Re: (Score:2)
Re: (Score:2)
It's pretty easy to defend a practice that kills people while you know you are not going to be one of those people.
How do I know that? Why am I less likely to die than anyone else?
Re: (Score:2)
Re: (Score:2)
That sure worked well for elevator operators.
Re: (Score:2)
Excellent (Score:2)
I am happy to see this, it should allow them to get more and better data in their limited area before rolling out on a larger scale. And the fact that anyone has done it should give more confidence to others to try as well. I had hoped that we'd have seen this 5 years ago, but better late than never.
Train safety operators (Score:2)
"Some of these will be "rider only," while other vehicles will still have train safety operators behind the wheel."
That would help if they were sensible self-driving vehicles, AKA trains.
Nice work, "editors"
Re: (Score:3)
That would help if they were sensible self-driving vehicles, AKA trains.
There is no train track from my house to the grocery store, nor will there ever be.
Re: (Score:2)
"There is no train track from my house to the grocery store, nor will there ever be."
You probably wouldn't like a full-scale track next to your house, but a nice PRT might be convenient. If you live in the boonies, though, you're last.
How good could this be? (Score:1)
Re: (Score:2)
I see it's almost time.. (Score:2)
Double your fun. (Score:1)