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

Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

techcrunch.com

Post reply on HN
Print view View original

191–200 of 456 posts

Re: Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

#191

Post by GloamingNiblets »

crazygringo wrote: post #173
GloamingNiblets wrote: post #118

Earlier quoted context omitted.

I don't have any specific knowledge about Waymo's stack, but I can confidently say Waymo's reaction time is likely poorer than an attentive human. By the time sensor data makes it through the perception stack, prediction/planning stack, and back to the controls stack, you're likely looking at >500ms. Waymos have the advantage of consistency though (they never text and drive).

What gives you that confidence? You're quite wrong. It tends to be more like 100–200 ms, which is generally significantly faster than a human's reaction. People have lots of fears about self-driving cars, but their reaction time shouldn't be on the list.

The better part of a decade as a SWE at another AV company. In practice the latency is a not a concern, I was just sharing some trivia.

Re: Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

#192

Post by ang_cire »

cynicalsecurity wrote: post #142
ericmcer wrote: post #60

Earlier quoted context omitted.

It must be interesting being an Uber driver right now and literally watching the robots that will replace you driving around with you. This has been a 15+ year process and will probably take a few more years. I don't feel too bad if they didn't manage to pivot in that time period.

> It must be interesting being an Uber driver right now and literally watching the robots that will replace you driving around with you. You mean just like programmers watching AI replacing them?

AI doesn't replace programmers, it's used by programmers for efficiency.

Waymo is most definitely not being used by taxi or rideshare drivers to be more efficient.

Re: Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

#193

Post by JumpCrisscross »

kjkjadksj wrote: post #187
vasusen wrote: post #92

I am really excited for this. Once going home with my family via Uber in SFO we realized on the freeway that our driver was high and driving at 80-85 mph. It was a really scary experience and I couldn’t do much about it in the moment.

How could you tell they were high?

> How could you tell they were high?

In New York it's not too difficult. Fidgitiness, twitchiness, rambling series of non sequiturs that make even my ADD brain rattle. Screaming at traffic and running on the margin one second and then asking me if I know that the archangel who visited Muhammed was actually a demon the next. (I'm not Muslim. The conversation wasn't addressed to anyone in the vehicle.)

Like, I guess I can't say they're taking too much of a substance. But if they aren't, they're taking too little.

Re: Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

#195

Post by jwagenet »

sib wrote: post #164
jeffbee wrote: post #85

Earlier quoted context omitted.

My hot take is that anyone who would take a taxi from JFK to Manhattan, along the most well-served transit corridor on the continent, is probably a psycho and we shouldn't ask for their input on transportation topics.

There are many, many, many airports to which it is easier to travel via public transit from their associated city than it is from Manhattan to JFK. For example, all of these global-top-25 airports have single-train access: London Heathrow (LHR) Tokyo Haneda (HND) Amsterdam Schiphol (AMS) Paris-Charles de Gaulle (CDG) Frankfurt (FRA) Dubai (DXB) Seoul Incheon (ICN) Guangzhou (CAN) Shanghai Pudong (PVG) New Delhi (DEL)…

I don't get the gripe. AirTrain gets you to A,E,J,Z, and LIRR, all of which get you to "Manhattan" or a significant number of intermediate destinations in about an hour. LGA is far worse.

Re: Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

#196

Post by creer »

NullHypothesist wrote: post #2

This is a huge sign of confidence that they think they can do this safely and at scale... Freeways might appear "easy" on the surface, but there are all sorts of long tail edge-cases that make them insanely tricky to do confidently without a driver. This will unlock a lot for them with all of the smaller US cities (where highways are essential) they've announced plans for over the next year or so.

Isn't really the main problem, the Waymo "let's just stop right here" current failure mode? Which really is not ideal on city streets either. Hopefully they have been working on solving that.

Re: Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

#197

Post by estearum »

CPLX wrote: post #135
QuercusMax wrote: post #117

Earlier quoted context omitted.

"People should break traffic laws" is a very strange position to take.

I'm sure it is in places that are dominated by strip malls and tract housing. Here in New York City, we have a different approach altogether. I find it much simpler and more straightforward and easy to understand. You always know exactly what another car is about to do. They are going to try to get in front of you and try to get where they are going, while not caring if that helps you go where you're going. I never h…

I live in New York City: no.

If everyone drove on the shoulder like a few assholes do, we just wouldn't have shoulders.

This is extremely silly.

Re: Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

#198

Post by GloamingNiblets »

embedding-shape wrote: post #172
GloamingNiblets wrote: post #118

Earlier quoted context omitted.

I don't have any specific knowledge about Waymo's stack, but I can confidently say Waymo's reaction time is likely poorer than an attentive human. By the time sensor data makes it through the perception stack, prediction/planning stack, and back to the controls stack, you're likely looking at >500ms. Waymos have the advantage of consistency though (they never text and drive).

> but I can confidently say [...] you're likely looking at >500ms That sounds outrageous if true. Very strange to acknowledge you don't actually have any specific knowledge about this thing before doing a grand claim, not just "confidently", but also label it as such. They've been publishing some stuff around latency ( https://waymo.com/search?q=latency ) but I'm not finding any concrete numbers, but I'd be very surp…

My experience is from another prominent AV company; I do not have Waymo insider knowledge.

Re: Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

#199

Post by bcrosby95 »

throwup238 wrote: post #184
mikestew wrote: post #155

Earlier quoted context omitted.

It’s more complicated than that because several (most?) states have contradictory laws about impeding traffic. No they don’t, you’ve misinterpreted what was written. “Not impeding traffic” is not codified as “exceed the speed limit if everyone else is, or get a ticket”. Or perhaps you have a documented counter-example.

I don’t know what you mean by “documented” but here is Georgia: > No person shall drive a motor vehicle at such a slow speed as to impede the normal and reasonable movement of traffic, except when reduced speed is necessary for safe operation. [1] Versus California: > No person shall drive upon a highway at such a slow speed as to impede or block the normal and reasonable movement of traffic unless the reduced speed…

> I don’t know what you mean by “documented” but here is Georgia:

Georgia isn't going to punish you for going the speed limit in the right lane, they passed that law recently and called it the 'slow poke law'.

> On the other hand “in compliance with law” is it’s own barrel of monkeys because it doesn’t specify priority.

It really isn't.

Re: Waymo robotaxis are now giving rides on freeways in LA, SF and Phoenix

#200

Post by JumpCrisscross »

embedding-shape wrote: post #22
NullHypothesist wrote: post #2

This is a huge sign of confidence that they think they can do this safely and at scale... Freeways might appear "easy" on the surface, but there are all sorts of long tail edge-cases that make them insanely tricky to do confidently without a driver. This will unlock a lot for them with all of the smaller US cities (where highways are essential) they've announced plans for over the next year or so.

> Freeways might appear "easy" on the surface, but there are all sorts of long tail edge-cases that make them insanely tricky to do confidently without a driver Maybe my memory is failing me, but I seem to remember people saying the exact opposite here on HN when Tesla first announced/showed off their "self-driving but not really self-driving" features, saying it'll be very easy to get working on the highways, but th…

> remember people saying the exact opposite

It was a common but bad hypothesis.

"If you had asked me in 2018, when I first started working in the AV industry, I would’ve bet that driverless trucks would be the first vehicle type to achieve a million-mile driverless deployment. Aurora even pivoted their entire company to trucking in 2020, believing it to be easier than city driving.

...

Stopping in lane becomes much more dangerous with the possibility of a rear-end collision at high speed. All stopping should be planned well in advance, ideally exiting at the next ramp, or at least driving to the closest shoulder with enough room to park.

This greatly increases the scope of edge cases that need to be handled autonomously and at freeway speeds.

...

The features that make freeways simpler — controlled access, no intersections, one-way traffic — also make ‘interesting’ events more rare. This is a double-edged sword. While the simpler environment reduces the number of software features to be developed, it also increases the iteration time and cost.

During development, ‘interesting’ events are needed to train data-hungry ML models. For validation, each new software version to be qualified for driverless operation needs to encounter a minimum number of ‘interesting’ events before comparisons to a human safety level can have statistical significance. Overall, iteration becomes more expensive when it takes more vehicle-hours to collect each event.”

https://kevinchen.co/blog/autonomous-trucking-harder-than-ri...

Post reply on HN