Waymo

[AlanSubie4Life]

80% of the time it could have just blown the stop sign and made the left without any issues.
I don't think pushing through while cars are stopping for the light is a failure. The failure was not pulling all the way out. The cars even stop to let it through and yet it sits blocking the lane! Failure.
View attachment 898845

Yes, there are several correct approaches but this one was not it. How hard is it to figure out people are stopping for you? Just look at the driver! And of course them being…stopped.

If you’re going to pull out in the street (rather than just wait for a time when crossing is easy) you have to commit to it and assert your right of way (that you don’t have but are taking once it is given). I don’t really see that that was the plan here.

I prefer waiting because it is possible people won’t give their right of way, and then you are just left sitting halfway into the street. And you never know what could happen. Someone could be driving along in the empty lane watching a YouTube video, and I prefer a rear collision rather than a t-bone in that case.

 

[enemji]

If the city had nothing but Waymo vehicles with the current software & hardware, I wonder how it would have managed. Thoughts?

 

[Doggydogworld]

Remember, last year, Waymo switched to an all new Planner. The old Planner was hardcoded to be overly cautious all the time. The new Planner uses NN.

Both planners use NNs as well as hand code. I'm sure the new one has more sophisticated NNs.

Patience was in order. Trams move.

And in this case Waymo ended up waiting until after the tram moved, anyway.

But there was not really any reason to block one lane. It didn't improve visibility....

Of course it improved visibility. Especially for the top mounted sensors. Maybe it didn't improve visibility enough to proceed. Or maybe Waymo saw the black Prius coming from the right and decided to wait instead of blocking the path of the cars on the left that were giving it space.

Waiting for the Prius that was ~7 seconds away is a very cautious move, though. I agree it doesn't make sense to aggressively pull into one lane of cross traffic then suddenly turn into Cautious Carl. On the other hand, there may have been a car even closer than the Prius that ended up turning or something.

It should have gone out and blocked both lanes, then waited for a gap (which it could assume would appear when the the traffic light to the right changed).

How would that light changing from red to green create a gap? The opposite is what actually happened.

 

[Daniel in SD]

How would that light changing from red to green create a gap? The opposite is what actually happened.

The lights are red briefly in both directions before it turns green.

 

[EVNow]

I guess people will see what they want to see. I call it a success. Waymo moved into the lane in order to see past the occlusion, but was safe and made the turn with confidence.

Lots of times when Curt said the ULT was a success still @AlanSubie4Life insisted they were failures.

But I've to say, IMO, this was handled worse than what Tesla does.

Anyone local who can take your Tesla there and check it out ?

 

[diplomat33]

But I've to say, IMO, this was handled worse than what Tesla does.

You think Waymo handled that UPL worse than what FSD Beta would have done? Are you crazy? There is zero chance FSD Beta would handle that situation even 10% as good as Waymo. The fact is that Waymo handled that UPL incredibly well.

 

[EVNow]

You think Waymo handled that UPL worse than what FSD Beta would have done? Are you crazy? There is zero chance FSD Beta would handle that situation even 10% as good as Waymo. The fact is that Waymo handled that UPL incredibly well.

Waymo would never even try CULTs.

 

[diplomat33]

Waymo would never even try CULTs.

Wrong. IMO, Wayno would not only try CULTs but would pass with flying colors. CULTs would be a piece of cake for Waymo.

 

[EVNow]

Wrong. IMO, Wayno would not only try CULTs but would pass with flying colors. CULTs would be a piece of cake for Waymo.

LOL. Dude, you really jump from one camp to another, don't you ?!

 

[Bladerskb]

But I've to say, IMO, this was handled worse than what Tesla does.

Waymo would never even try CULTs.

If Tesla is soooo good and waaaay better than Waymo as you and other Tesla fans repeat everyday.
WHY haven't they launched driverless? They are losing out on tens and hundreds of billions of dollars each year by not launching. While they have the superior technology.

Especially when they are 5-10 years ahead depending on which Tesla fan you ask.
With their "Unsurmountable lead". Why in the world do they not have a driverless paid robotaxi yet?

Whenever i ask a Tesla fan that, I get an excuse in response or something that doesn't make logical sense in response.

Like SolvingTheMoneyProblem says Tesla will launch a driverless robotaxi in a city somewhere in the US like phoenix or SF in 2024. Wait you mean that Tesla who is 10 years ahead according to him would be doing what Waymo was doing 5 years ago in 2024? LOL that guy's brain is SEARED!

So then I ask you, If Tesla is soooo good and waaaay better than Waymo as you claim WHY haven't they launched driverless service anywhere? Why haven't they released any REAL data on FSD?

 

[diplomat33]

LOL. Dude, you really jump from one camp to another, don't you ?!

Not at all. I only jumped once after I realized Elon's claims about FSD were BS. I have not jumped anywhere since. I am in the camp of real autonomous driving. That's where I am now and where I will stay.

 

[EVNow]

If Tesla is soooo good and waaaay better than Waymo as you and other Tesla fans repeat everyday.

Just the same way GM is not launching Ultra-cruise.

Do you work for GM ?

We are just talking about CULTs ... not generic geo-fenced $50k sensored extreme money losing ventures.

 

[Bladerskb]

Just the same way GM is not launching Ultra-cruise.

Ultra Cruise is almost certainly not launching (consumer availability) in 2023. Doesn't matter as the incompetent management at GM who nailed Supercruise to a single car model for years is taking the same disastrous path with Ultra cruise by handcuffing it to a 300k car model.

Nevertheless this is pure deflection, my question has nothing to do with Ultra-cruise. You still have yet to answer me.

Do you work for GM ?

I rather not talk about my employment as Tesla fans/community has shown a history of trying to use it against people to get them fired.

We are just talking about CULTs ... not generic geo-fenced $50k sensored extreme money losing ventures.

Excuse 101

We already established that Tesla is 5–10 years ahead of Waymo.
We already established that Tesla's hardware/sensors is wayyyyyyyyyyyy better than Waymo and also wayyyyyy cheaper.
We already established that Tesla makes huge profits on their FSD technology.
We already established that Waymo is losing money because the cost of their engineering team and operations outweighs the revenue they are getting.

That's not what's in question.
You claim, like other Tesla fans, that Tesla is ahead or better.

So the simple answer I'm asking is: WHY are they not DRIVERLESS ANYWHERE?!
It's a simple question that requires a simple answer.

But all i get are deflections.. "Lmao, glorified AEB that's geofenced to train tracks that cost 50k."
But there's a reason why none of you can answer the question without giving an illogical answer.

 

[EVNow]

I rather not talk about my employment as Tesla fans/community has shown a history of trying to use it against people to get them fired.

But since you apparently work for a competitor, there is an obvious conflict of interest that you should disclose.

When I worked in MSFT, I put that in my sig so that people knew where I was coming from (in forums that talked about MSFT & competitor products). Here I've always included in my sig that I'm long on TSLA. Thats called ethics.

You don't have to say who you work for, but a little "I work for a competitor of Tesla" in the signature would be appreciated.

Excuse 101

When you jump into a conversation it is important for you to understand the context. Go back and check the context.

 

[Doggydogworld]

Like SolvingTheMoneyProblem says Tesla will launch a driverless robotaxi in a city somewhere in the US like phoenix or SF in 2024.

One or two cities? What an idiot. We all know Tesla will launch Everything, Everywhere All At Once.

 

[Bladerskb]

Go back and check the context.

The question is still valid and will/should be asked every time someone says "Tesla FSD can do X better".

 

[diplomat33]

https://twitter.com/x/status/1617766603231498245

This gives us a hint of where the Waymo service area in LA might be when they launch.

This is ~4 miles from LAX airport. So I think this is a strong hint that Waymo is planning to launch their LA ride-hailing service around the airport, similar to how the ride-hailing in downtown Phoenix services the Sky Harbor Train for the Phoenix airport. I think it is smart for Waymo to focus on airports. When Waymo does launch driverless ride-hailing to and from LAX that will be a great business model.

 

[enemji]

Do you work for GM ?

we neednot to be asking such information on a public forum. Enjoy the anonymity and freedom to discuss without being locked into a perceived stereotype

 

[Tam]

Although Autonomous Vehicles can be very good at collision avoidance but it can cause traffic back up when it didn't know how to proceed at 8:47 a.m. today on 19th Avenue and Ulloa Street in the Inner Sunset:

https://twitter.com/x/status/1617927201542000646

 

[cwerdna]

Driverless Waymo brings SF traffic to a halt during rush hour

The car was seemingly confounded by “a very complex and busy intersection.”

www.sfgate.com

is on the above.