Waymo says they're not doing that. Also it would be crazy unsafe. The cellular network is nowhere near reliable enough to remotely control a car.
Presumably. That's what most people do when their car stops working.
-
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
Waymo will allow more people to ride in its fully driverless vehicles in Phoenix
So they insist the article says. I'm not really saying drivers would remote in and 100 percent control the car in a high speed situation where latency from the cell connection would contribute. I'm talking more about the car slowing down, becoming uncertain in the path forward, and remote operators watching the feed suggesting a route forward to the car if they determine safe to proceed. Or, approving a route the car wants to take but has less confidence in. The cars sensors would still be active and the car would in essence be steering itself (maybe just off the HD maps path it would have normally taken).
Not really a great comparison about getting out of car. True if my car completely died I would get out. But in this sense car is not dead, just doesn't know the path forward. So far in life, I've yet to get to a point where I could not navigate and exited the car.
Either way impressive, no doubt about it.
"These driverless vehicles aren’t totally alone in the wilderness. Waymo has a team of remote employees who watch the real-time feeds of each vehicle’s eight cameras and can help, with the push of a button, if the software runs into a difficult spot and needs a human eye to figure out what’s going on. But Waymo insists these remote operators won’t be “joy-sticking” the cars, which are outfitted with a bevy of cameras and sensors that help it “see” its surroundings."
- theverge.com
Undoubtedly Waymo is doing a significant amount of work designing the remote operator/car interface.
- theverge.com
Undoubtedly Waymo is doing a significant amount of work designing the remote operator/car interface.
diplomat33
Average guy who loves autonomous vehicles
They are not. Waymo is actively testing in cities with poor weather conditions.
In 2017: Waymo started testing in Michigan to deal with extreme cold and snowy conditions.
Waymo starts testing in Michigan to master snow and ice
In 2019: Waymo started testing in Florida to deal with heavy rain conditions.
Waymo is rain-testing its self-driving cars in Florida
Maybe you are overlooking the most key factor here... sure occasionally Waymo will request input from remote center.. but there are is no human continuously supervising and ready to step in when necessary. If Waymo had no remote input, they could still drive safely across town with no driver... however, on the rare occasion if they get stuck, they stay stuck or pull over... this leads to a bad user experience.. not a safety issue. If a driverless Waymo requests input once every thousand miles, then you can have a small team of human assistants supporting a fleet of several thousand driverless cars.
This is the biggest misconception about Waymo... that their technology in only works in Pheonix Arizona
Also, FYI, for people on this thread I know someone plans to record several long uninterrupted driverless Waymo videos in complicated areas and upload them starting next week. I am also working on a way to get approved to Waymo One to travel there later this year for fun.
diplomat33
Average guy who loves autonomous vehicles
@NHK X Waymo tweeted that "driverless" means no remote operators. So Waymo has removed remote operators too for the driverless rides.
That is pretty dang impressive now that details are clear. I got nothing against Waymo. I go to Tucson/Phoenix pretty often, will try to get into the service one of these days.
Will be interesting to see how they scale in the future given differences in approach with Tesla (and same for Tesla).
I guess as Diplomat has shown the Tweet from Waymo below, zero remote input.
diplomat33
Average guy who loves autonomous vehicles
If you get in, please share your experience. I always love hearing about the Waymo driverless experience.
Yes, it will be interesting. Waymo's approach seems to be get the FSD really good first and then expand to new areas. Right now, Phoenix seems to be their testbed for ride-hailing. They are using Phoenix not just to improve the autonomous driving but also to improve the business and customer service aspect of ride-hailing. But Waymo definitely has big plans for scaling up.
Here is what we know so far about Waymo scaling efforts:
1) Expand the driverless area in Phoenix from 50 sq. mi. to 100 sq. mi. Maybe more after that?
2) Continue to improve autonomous driving and ride-hailing experience.
3) Launch Waymo One in LA or SF. Maybe other cities after that?
4) Waymo has an order for 20,000 Jaguar I-Pace's that they can tap into when they need to increase their fleet size. Waymo currently has 600 robotaxis. An extra 20,000 would be a significant increase in fleet size.
5) Waymo already has partnerships with Chrysler, Jaguar, and Renault-Nissan to provide them with vehicles. Waymo also plans to go with 100% electric cars too. So they already have 3 OEM partnerships in place to get a 100% electric robotaxi fleet.
5) Waymo also has plans to lease robotaxis as a yearly subscription direct to consumers.
I feel like Waymo has a lot of potential to scale up fast. They obviously have good FSD and they have the partnerships in place to dramatically increase the size of their fleet.
Yep, I am going to bookmark that tweet. though it doesn't say 'zero remote input', it just says 'no safety drivers'. Occasional high level input != safety driver
That's not what it says. It says no remote driver.
This is a remote driver. Waymo is not doing this.
Yes. Would be cool to see the insider view of how they operate, the view they have and what options they can exercise for the worst case scenarios. If it truly is just a simple screen they monitor with no input would be super impressive.
Possible, of course, but safe? I don't see how that could possibly be as safe as having a driver in the car.
I'm not sure they're using a cell phone network for that demo. They say it's an office a few blocks away.
They are only testing in cities Why not expand to rural areas between cities It would be nice to take one of these cars from a Rural town to a major city when the temp is 35 degrees below zero
Long range commute is one of the congestion problems that makes FSD problematic for cities. A four hour round trip commute becomes acceptable for some people. Sleep on the inbound. Finish work on the outbound. Tell your car to charge and then find a place to park while you are in the office. Maybe it can even go pick up lunch and save the three dollar delivery fee.
This optimal individual behavior can't work in currently congested areas.
Similar threads
- Article
