Waymo Makes History: First Fully Self Driving Car With No Driver

[Tam]

...And as has been pointed out, Tesla is operating in CA...

Tesla was able to test its Roadster in 2014 inside the Arctic Circle which is in another continent far from CA!

There's no technological nor legal reason that Tesla cannot test its ability in those 3 states that allow driverless cars (not just testing driverless cars.)

 

[Tam]

...I couldn't see a point where the Waymo cars displayed more than 25 mph on their speedometers...

I don't mean the speed for the video demo.

I mean the Koala driverless car maximum rated speed was 25 MPH while current Chrysler Pacifica is designed for highway speed as well.

 

[Tam]

...It’s a demo in a 100 square mile area....

Yes. Geofencing is a limitation but it doesn't mean the system is useless.

A company like Tesloop that has only one specific route--Los Angeles and Las Vegas--could take advantage of such limitation.

 

[malcolm]

Yes. Geofencing is a limitation but it doesn't mean the system is useless

How did you get from demo to useless?

It's not useless at all. The video shows residents who have signed up for the system being taxied slowly around their quiet suburban neighbourhood.

According to the evidence in the video it looks stable in low levels of traffic and at modest speeds.

But it does remind me of Blue Origin's claim of reusability.

Which was another example of a big claim extrapolated from a limited demo.

 

[AnxietyRanger]

There's no technological nor legal reason that Tesla cannot test its ability in those 3 states that allow driverless cars (not just testing driverless cars.)

I bet you there is.

 

[AnxietyRanger]

How did you get from demo to useless?

It's not useless at all. The video shows residents who have signed up for the system being taxied slowly around their quiet suburban neighbourhood.

According to the evidence in the video it looks stable in low levels of traffic and at modest speeds.

But it does remind me of Blue Origin's claim of reusability.

Which was another example of a big claim extrapolated from a limited demo.

That's no demo, though. It is a pilot project. A pilot that includes production use in limited scope.

Demo is just a demo.

 

[Tam]

...It is a pilot project..

Early Rider Program for the public in some cities (Phoenix metropolitan area, including Chandler, Gilbert, Mesa and Tempe, AZ) has been running for the past 7 months but with Waymo engineer in the driver seat.

With this announcement, the same program is now moving up to an empty driver seat:

"Over the next few months, we’ll be inviting members of the public to take trips in our fully self-driving vehicles. Participants in our early rider program will be amongst the first to experience these fully self-driving rides, using our vehicles to commute to work, take the kids to school, or get home from a night out. When fully self-driving vehicles become part of people’s everyday routine, we can move closer to our goal of making transportation safe and easy for everyone."

 

[Tam]

...What Waymo has done here is impressive. To suggest, however, that they beat Tesla somehow is disingenuous.

No one has reached the finish line just yet so declaring win or lose at this time is still premature.

So, is Tesla ahead?

Tesla has not activated Self-Driving Capability yet and all cameras are not even activated either.

I would say Tesla is behind at least in this area while Waymo has activate its many sensors already.

5 yellow LIDARs
4 dark blue RADARs
1 green Vision System
1 bright blue Supplemental Sensors

Tesla has not announced that it would allow a driver seat to be empty but Waymo does.

I would say that Tesla is behind in these areas.

What's Tesla ahead in?

Whatever Tesla is behind, its system still can work in different places from freeways to local streets.

That is not the case with Waymo. Its operation only works at specific places that have been prepared and mapped for (geofenced).

Can Tesla pull ahead?

Since Tesla can work in familiar as well as new locations, once Tesla can figure out its driverless capability, it can work globally while Waymo might still have to catch up with mapping and preparing for a prospective locations/routes.

 

[CameronB]

Taken from the Waymo FAQ:

“Our goal is to develop fully self-driving vehicles that require no intervention, though as part of this early trial, there will be a test driver in each vehicle monitoring the rides at all times.“

So in AZ, the “driver” can sit in any seat in the car? So in that video, the “WAYMO 02” car where there was only one occupant...that was a Waymo employee driver? And in the other two Waymo cars, one of those two people would be a “backseat driver”? Did I get that right?

 

[Tam]

...a Waymo employee driver...

I understand the same way as you do.

In this phase, the driver seat is kept empty. At least a Waymo employee will be present in the car in the beginning. Eventually, they may decide to drop that all together.

"But instead of being in the front seat, that employee will likely sit behind the driver’s seat. (Waymo says that won’t always be the case, though.)"

 

[AnxietyRanger]

I find this thread puzzling. I mean, I get the disbelief when we discuss, say, Audi, or even GM's autonomous work. Perhaps even Delphi, who did the coast-to-cost on an Audi excluding some urban parts. We can go back and forth on those.

But Waymo? I mean, Waymo is to autonomous cars what Tesla is to battery electric vehicles. Right?

These guys are the pioneers of urban autonomous. Even George Hotz believes they will be first-to-market. We can argue their impact (Hotz certainly does) as a commercial/limited fleet-based non-car-maker solution, but I don't think we can argue their leadership in autonomous driving.

Whereas almost everyone else (Tesla included) has been solving only the easy part - motorways - so far in public products and public testing, Waymo has been doing the hard part - urban driving - for soon a decade.

There is a staggering double standard going on in this thread. Tesla's theoretical future features are compared to Waymo's actual, public features. Just because Waymo chooses to limit their shipping Level 4 product to a certain geofenced area at first, doesn't mean they aren't capable of doing it elsewhere.

Tesla does not do Level 4 anywhere yet, Waymo does once this operation goes live. That's huge.

I am pretty sure if AP2 ever gets to Level 4, it too will start out as a geofenced product.

 

[Tam]

...George Hotz...

He favors Tesla and thinks Waymo is doing the wrong way which will delay the effort further. Waymo autonomous operation won't work in an unfamiliar location. That location must be first prepared as he described below:

"Tesla understands more about what we're doing, but let's look at Waymo. So Waymo, when they want to figure out how to build a self-driving car, they sit four engineers down in a room and they talk about, "Okay, well we come upon a stop sign, we know that we should stop and we know that we should stop this distance, and wait, can you get the DMV handbook out? Let me see. Okay, we have to signal 15 meters—," That's not what driving is. This is the same failure of computer vision for many, many years."

 

[Canuck]

Waymo Makes History: First Fully Self Driving Car With No Driver

Well it's a good thing they didn't make history with the first self driving car with a driver.

 

[AnxietyRanger]

He favors Tesla and thinks Waymo is doing the wrong way which will delay the effort further. Waymo autonomous operation won't work in an unfamiliar location. That location must be first prepared as he described below:

"Tesla understands more about what we're doing, but let's look at Waymo. So Waymo, when they want to figure out how to build a self-driving car, they sit four engineers down in a room and they talk about, "Okay, well we come upon a stop sign, we know that we should stop and we know that we should stop this distance, and wait, can you get the DMV handbook out? Let me see. Okay, we have to signal 15 meters—," That's not what driving is. This is the same failure of computer vision for many, many years."

In his recent interview Hotz does say he believes Waymo will deliver on autonomous first. I think even he must recognize their massive lead.

It is true he thinks they will have a very limited impact due to scale and, of course Hotz is a big believer in self-driving being solved through deep learning.

Interesting to see who implements NN driving car first. So far Tesla and Comma.AI certainly have not used NN to drive either in their public versions.

The thing is, even if we assume what Tesla and Comma.AI is doing is faster than what Waymo has been doing, Waymo has been doing it - in some form or another - since 2005.

That's a lot of catching up to do. The same thing with MobilEye. Sure, some of their lead may evaporate due to the fast progress in deep learning, but that doesn't automatically translate into bypassing them.

In the meanwhile, Waymo is driving kids around in urban environments autonomously while Tesla and Comma.AI only have motorway driver's aids in public.

 

[S4WRXTTCS]

What's the obsession with who's ahead?

Waymo, and Tesla are two VERY different things. One is meant as a ride hailing company, and the other is a product you can buy. They are two wildly different approaches that will likely meet at some point, but not for awhile.

Tesla will add capabilities through HW/SW upgrades, and Waymo will slowly expand their working circle. Hopefully Waymo will license their technology to manufactures so people can actually buy the cars. But, that might never happen.

I would say the biggest rival to Tesla is GM with the self driving Bolt. But, Bolt buyers right now can't even get TACC.

 

[malcolm]

I find this thread puzzling....

Waymo is to autonomous cars what Tesla is to battery electric vehicles....

These guys are the pioneers of urban autonomous....

Whereas almost everyone else (Tesla included) has been solving only the easy part - motorways - .........Waymo has been doing the hard part - urban driving - for soon a decade.

There is a staggering double standard going on in this thread......

Tesla does not do Level 4 anywhere yet, Waymo does once this operation goes live.....

Way-mo' miles to go before we sleep (on the back seat)

You're comparison with EV development is an interesting one.

In the same way that early EVs were considered Punishment Vehicles - low range, low speed etc - this present iteration feels like Punishment Autonomy.

Yes, the cars look a bit better - but there's still that strong first-responders-on-holiday look.

Which is great if you happen to be a first responder or you're running commercial deliveries etc.

Yet the average speeds have only increased by 10-15mph.

This was ten years ago:

DARPA Grand Challenge - Wikipedia

Are these Waymo vehicles the "Tesla Roadster moment" for Autonomy?

Or to put it another way: "Who's current take on Autonomy feels the least like some form of punishment?"

 

[AnxietyRanger]

Or to put it another way: "Who's current take on Autonomy feels the least like some form of punishment?"

There certainly are points that are made against Waymo, many of the jobsesquely made by Elon and repeated by those who believe similary. It think we can summarize than something like this:

1) Lidar is unnecessary and too expensive
2) Waymo is niche as a ride-hailing operation, lacks consumer fleet
3) Waymo relies too little on deep learning, slow traditional approach
4) Waymo is ugly, i.e. your point, see 1)

Certainly, if Tesla can bypass Waymo's progress through deep-learning, fleet-learning and integrated consumer-available sensors and put this all out as Tesla Network on every Model 3 sold, that would be a Waymo killer due to volume. If you believe that, I can certainly underestand why you might tend to ignore Waymo.

The thing that baffles me is, does anyone believe that, though? I mean we can believe in Tesla's vision (no pun intended), but does anyone actually believe Tesla will come out with Level 4 software update to their cars in the next few years? A go-to-sleep AP2 update within 2017-2019? All the while Waymo will be expanding on their lead in more and more areas, probably in an exponential fashion... At the very least, that would take an expontential leap technologically from Tesla, while all Waymo technologically needs now is evolution and can look forward to growing their presence.

The looks question can probably be improved much more and much more quickly than driving software can be made - that is not a big problem solve, if it is a problem. And the hardware will only get cheaper and better.

But at the same time, people are missing the real paradigm shift here. Sure, looks matter when you are talking about a limited driver's aid selling a traditional car, but if I could get a car that can reliably drive me without a driver, I would not care at all what it looks like - any more than I care what a taxi looks like as long as it is comfortable inside. Safety matters too, I actually think that "ambulance suite" might actually appeal to some people's safety requirements - it is, for sure, far more redundant than Tesla's is...

 

[Shock-On-T]

Yes, yes it is. Guarantee that they ran that exact route multiple times before making the video.

If they were confident in it, they would let journalists give a try instead of just showing a canned demo.


Note, I'm not saying its not further along than Tesla, or that it isn't close to being production ready. Just saying that video is no better than what Tesla showed last year.

Further Note. They give no indication of how much all of their hardware costs. If they need $300k worth of gear on each car to make it work...

A taxi license used to cost $300k.

 

[Shock-On-T]

What's the obsession with who's ahead?

Waymo, and Tesla are two VERY different things. One is meant as a ride hailing company, and the other is a product you can buy. They are two wildly different approaches that will likely meet at some point, but not for awhile.

Tesla will add capabilities through HW/SW upgrades, and Waymo will slowly expand their working circle. Hopefully Waymo will license their technology to manufactures so people can actually buy the cars. But, that might never happen.

I would say the biggest rival to Tesla is GM with the self driving Bolt. But, Bolt buyers right now can't even get TACC.

No, the biggest rival to Tesla is whichever giant car company Waymo licenses its product to.

 

[malcolm]

Deja vu?

Blue Origin lands a rocket stage

Is it Tesla = Blue Origin or is it Waymo = Blue Origin?