Interesting time to stop the clip if there really wasn't a driver.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Waymo
- Thread starter Daniel in SD
- Start date
diplomat33
Average guy who loves autonomous vehicles
Waymo just posted a blog on how they handle rain and fog:
blog.waymo.com
A fog blog: Understanding a challenge inherent to driving in San Francisco
If you check the weather for San Francisco, there's a one third chance you’ll see clouds or fog in the forecast, according to the National Oceanic and Atmospheric Administration (NOAA). To ensure Waymo’s Trusted Testers can safely get where they are going, whether it’s to work in the misty...
blog.waymo.com
I see no mentioned of planned driverless Waymo trucks in press coverage of their autonomous truck efforts in Arizona and Texas as recently as August.
I’m guessing there really was a driver or two in the truck but they couldn’t be seen due to the angle of the camera.
diplomat33
Average guy who loves autonomous vehicles
The truck in the video looks just like the Waymo Via truck seen in this link:
Waymo Via
In the posted video, you can see the spinning Lidar units installed near the mirrors. I think it's a real thing whether in testing or actual delivery usage.
That said, I'm a little surprised they don't have a safety driver sitting in front of the steering wheel, and the linked video stops before you can verify there was really no one there.
Waymo Via
In the posted video, you can see the spinning Lidar units installed near the mirrors. I think it's a real thing whether in testing or actual delivery usage.
That said, I'm a little surprised they don't have a safety driver sitting in front of the steering wheel, and the linked video stops before you can verify there was really no one there.
I would say yes, but also that such an event should become almost unheard of with such advanced vehicles.The issues that lead to trucks needing to use those runaway ramps have to do with inadequate prior maintenance and/or poor operator technique (failure to downshift and engage, or using the friction brakes in a continuous application that severely overheats them).
Autonomous trucks will l presumably not make such operational mistakes in downhill braking, will undoubtedly be required to have a very formal maintenance and inspection regime, and will have advanced on-board diagnostics that measure and give early warning of under-performing brakes.
And that refers to conventional ICE trucks (not ice as in frozen water delivery, nor Ice Road Trucks as in the reality show
). Electric trucks will have strong regen braking capability to replace the engine braking of conventional trucks, and probably the ability to reverse oe re-phase the motor drive while in forward motion for incredibly powerful motor braking. Plus of course conventional friction brakes, and probably redundancy of all of the above. I ghink the real limit would be the tire traction limit, and that can be severely degraded by weather conditions or just excessive dirt/sand/mud/gravel on the road.So in my view, the old downhill runaway-truck scenario will become pretty much a thing of the past. OTOH a big truck is still a dangerous moving mass in any number of more general accident situations, most of which will not happen to have an emergency escape path waiting to be used.
Perhaps (just a thought) some sort of emergency system that actually drops anchor and scrapes or digs into the road surface to provide a (literally) last-ditch method of slowing the truck in a dire situation. Thinking more about this last rather destructive method, it would be especially important that the AI doesn't ever get confused and throw an unnecessary Phantom Braking event - because for once this would actually be the "Phantom Breaking" scenario of Tesla internet lore...
EVNow
Well-Known Member
diplomat33
Average guy who loves autonomous vehicles
Why is Waymo in even deeper trouble? The main issue that everybody has with autonomous driving, not just Waymo, is solving all the complex interactions that can happen between road users and making sure that your autonomous driving is safe enough in all those driving interactions. And Waymo can't scale until they do that. But that is the challenge of deployment that everyone has. Waymo is making very good and steady progress on that front. Obviously, it is a difficult challenge. Hence why nobody has deployed driverless FSD at scale yet.
EVNow
Well-Known Member
... then they are in deeper trouble (than I thought). Imagine if they had solved all the "interaction issues" and just needed mapping. Obviously that would have been a better situation for Waymo.
If they are taking this long to resolve interaction issues - may be they need to look at end-to-end NN. They definitely have the resources, so why not ?
Unlike Tesla (and Mobileye, Cruise) , they need to solve the robotaxi and deploy in a LOT of cities in large numbers to make any kind of money. That is why they are in deep trouble. 10 years and hardly a dime earned. Wow.
ps : This is also the reason why monopolies like Google shouldn't be allowed to exist. They can outcompete other companies in other industries by using monopoly money in one or two industries.
pps : I still see no link and assertions about "mapping is fully automated". Where are the links ?
You should be more worried about your boy Elon
EVNow
Well-Known Member
You got it all mixed up. I'm interested in getting L2+/L3 in my car - I've no illusions about actual "robotaxi" anytime soon.
You are just being disrespectful when commenting as usual. If you change your tone, may be we can have productive discussion.
EVNow
Well-Known Member
I'm in a generous mood today. Thanks to TSLA I'm retiring earlyYou're a optimistic person.
Doggydogworld
Active Member
How can it be safe to run driverless in 50 sq miles of Phoenix but not in the other 2950? And if the technology is still not quite ready to scale in easy Phoenix, why try to scale it in crazy-hard San Francisco?
There's only one reason to deploy in San Francisco: customers. (And investors, ha). The conclusion is obvious -- they couldn't attract customers in Phoenix.
Four years ago they believed Phoenix would work. That's why they ordered 82,000 vehicles, bought factory space in Detroit to outfit them and signed deals with Avis to manage/clean/charge them and Autonation to service/repair them. And why they custom-designed their own cost-reduced Gen 4 sensor set. You don't do any of that unless you're ready to deploy at scale. They had everything in place, except the customers. They built it, but no one came.
So now they're going where the customers are, even though SF requires much better technology. At this point they have no other choice.
Yep, business model trumps technology. It's 100x better to sell technology that doesn't work at $10k a pop to hundreds of thousands of eager fanboys than to develop working technology and have no idea how to sell it. Say what you want about Elon, he knows how to get people to write checks.
Who is Google "outcompeting" in self-driving? They're pouring money into R&D. Someone with a clue will probably take their work and run with it, like Jobs and Gates did to Xerox PARC.
EVNow
Well-Known Member
They scare away potential rivals. Like HRC did in '16 primary
In IBM's days this was called vapor ware.
Boris Sofmam, head of trucking at Waymo, says that they've fundamentally changed the hardware and software stack from what's deployed in Phoenix because the 4th generation system was not easily scalable.
EVNow
Well-Known Member
You mean he didn't admit / say, there were no customers? So all that stuff about buying so many Porsche (?) cars was a bluff?
What makes people believe whatever other company folks say but not Tesla ? I look at all company communication as marketing.
Similar threads
- Article
