Cruise Robotaxis Now Run All Day In San Francisco (electrek.co) 37
According to a recent Twitter post from Cruise CEO Kyle Vogt, the robotaxi service is now operating all day in San Francisco. The post says we will soon see Cruise "open up full operations in other cities," which may soon include Dallas, Texas, according to a recent job listing. From the report: According to a recent LinkedIn post from Cruise CEO Kyle Vogt, the robotaxi network is now running 24/7 rides across San Francisco, beginning with employees. As The Kilowatts points out on Twitter, nonemployees in the San Francisco area are still limited to about one-third of the city between f 10:00 p.m. and 5:30 a.m. In his post, Vogt said that in accordance with safety policies, around-the-clock public rides will roll out "very soon."
Cruise is a robotaxi startup founded in the San Francisco Bay area in 2013. In the last decade, the company (along with plenty of support from GM) has made tremendous progress in its home state of California, where it continues to try and expand. Services that began in San Francisco have since grown to Phoenix, Arizona, and, most recently, Austin, Texas. In February, the Cruise president, CEO, and cofounder, Kyle Vogt, shared that the company had surpassed one million miles driven without anyone behind the wheel. In many ways, the city by the bay has become a proving ground for Cruise's electric robotaxis, and its hilly, congested terrain will act as a testing site for yet another major milestone -- around-the-clock robotaxi operations.
Cruise is a robotaxi startup founded in the San Francisco Bay area in 2013. In the last decade, the company (along with plenty of support from GM) has made tremendous progress in its home state of California, where it continues to try and expand. Services that began in San Francisco have since grown to Phoenix, Arizona, and, most recently, Austin, Texas. In February, the Cruise president, CEO, and cofounder, Kyle Vogt, shared that the company had surpassed one million miles driven without anyone behind the wheel. In many ways, the city by the bay has become a proving ground for Cruise's electric robotaxis, and its hilly, congested terrain will act as a testing site for yet another major milestone -- around-the-clock robotaxi operations.
Re: (Score:3)
Dash cams can help, but a rear-end wreck is the fault of the person behind no matter what.
That seems unlikely if the dash cam shows the instigator's car backing up (and why wouldn't it?).
Re: (Score:3)
Re: (Score:2)
You can bet on that. They would not be allowed on the streets otherwise.
Re: (Score:2)
The poster above you is obviously not very smart. Crashes with self-driving vehicles that are not experimental where the self-driving vehicle was responsible are exceptionally rare _and_ the vehicle will have a detailed log available. Hence all this will do is getting some idiots convicted for attempted fraud.
Also "but a rear-end wreck is the fault of the person behind no matter what" is pretty wrong. This only holds if no intent was present and the car in front did not back up.
Re: (Score:2, Interesting)
Another common one is the person in front braking hard for no reason. In the UK at least it usually goes 50/50 if that happens, because the braking was unnecessary and the other car was following too closely.
It's become an issue lately because some cars with collision avoidance sometimes randomly brake.
Re: (Score:3)
Emergency braking has been available in cars for a long time, it's probably one of the easier things to automate (sensors detect distance to object in front and speed, brake accordingly) so even if you were able to get it to rear end you there are a few issues;
you would probably have to do it at high speeds putting yourself at risk, you could also be at fault due to an illegal lane change if you do it too close, that could be your fault, not the other persons, even if they hit you in the rear and 3rd the ca
Robots make consistent mistakes, huge advantage (Score:5, Interesting)
Robots make consistent mistakes, humans make random ones.
Consistent mistakes are fixable, random ones, less so.
Robots are better drivers not because they have better brains, but because they never make stupid mistakes.
Humans: start young, get old, get drunk, get tired, and get mad. We accept the first two, outlaw and punish the rest. Robots start inexperienced, blind, and ignorant of new activity. We FIX those and the robots do not repeat the problem.
Re: (Score:2)
Re: (Score:2)
You seem to be unaware that this is an iterative process that converges. Your claim is invalid.
Re: (Score:2)
I am not persuaded that the process converges. If it did, Windows would be bug-free by now after 30 years.
Re: (Score:2)
Yep, that nicely sums it up. One of the reasons why human drivers are a really bad idea as soon as automation becomes available.
Incidentally, that is also the reason why eventually robots will take all repetitive jobs.
Re: (Score:3)
I don't think that's true I spent a significant portion of my life chasing down (non-AI) intermittent bugs. With AI, it's a lot worse because you can't always recreate the exact scenario. Therefore the contrary to what you say is true with AI .. sometimes AI mistakes a turtle for an AK-47. That's why Kasparov said the best chess player is AI + average chess player. The average chess player can detect blunders. AI + expert is less effective because the expert second guesses the AI too often. AI tends to make
Re: (Score:2)
Yeah, but the tail is infinite. A human can encounter a totally new experience and figure out a reasonable way to deal with it. A robot less so.
I'm not against AI drivers. I just think that covering a decent chunk of the tail is a lot more work than anyone is really admitting.
Re: (Score:2)
Re: (Score:2)
There is a new science you may not have heard of yet which is able to systematize the types of mistakes made by humans, known by the catchy title "psychology."
It's often easier to find out why a human made a mistake than why a neural net made a mistake.
Re: (Score:2)
No, Robots are better because they don't get tired, they don't get distracted, they don't get mad. A road rager can come at them and they will continue to drive the sam
Re: (Score:2)
Being tired, distracted and mad are all subcategories of "Overstimulated with excess data". Robots do worse when given too much data as well. They slow down, and make mistakes. It is easier to tell them to ignore excess data, but that is not perfect.
Humans sample their environments MORE than robots do. Constant noise, temp, sight, touch, scent, samplings. The robots are the one that take too much time and quantimize their data into x times per second. Humans DO not take 2-8 seconds to react, we reac
One million miles (Score:3)
Is actually not a lot. A typical taxi does about 50,000 miles per year so you reckon 20 of them would get that.
Re: (Score:2)
That absolute number is irrelevant. What matters how well the coverage of to be expected situations is.
Re: (Score:2)
For comparison Waymo passed the 1 million mile mark back in 2015.
Seems like the reason they limited the operating hours is that they need a team back at base to monitor the taxis and intervene if necessary. All they have done is expand that to 24 hour coverage, and presumably simplified the task somewhat so that it doesn't require so much training or devops.
Re: (Score:2)
Is actually not a lot. A typical taxi does about 50,000 miles per year so you reckon 20 of them would get that.
It actually is plenty. What we're comparing against is not some fleet of professional drivers. What they are doing here is relating the cars to people in this statement saying that the number of driverless miles driven has surpassed the number of miles you are likely to drive in your lifetime, assuming you aren't a professional driver.
Do the Robottaxis run all night? (Score:1)
Do Dah, Do Dah
Do they video passenger behavior to prove damages? (Score:4, Interesting)
Re: (Score:2)
Re: (Score:2)
Re: Do they video passenger behavior to prove dama (Score:2)
Re: (Score:2)
Why is it that every story about any technology comes out someone always points out how it will be bad because American society is full of shitty feral pigs?
I mean it's everywhere. Based on my reading of Slashdot the only thing not stopping people pissing where they want is humans, not stopping everyone from shouting into their phones in a plane is lack of coverage and that there's nothing stopping anyone doing anything in a cinema and that all movies get rated by how much popcorn is thrown at you.
Is the pl
Re: Do they video passenger behavior to prove dama (Score:2)
I don't think that driverless taxis having to deal with drunken pissing, pugilism, procreation, etc, means that they're a bad idea. But, I do think it's a design consideration.
Hmmm... 'Behind the wheel' (Score:2)
It occurs to me that one could remotely control a car and not have anyone 'Behind The Wheel'.
Is there any data for any self driving cars maximum distance achieved 'Without Calling For Assistance' ?
"AVs About to Take Over!" - LOL (Score:2)
Ya'll remember 5-6 years ago when people were swearing that the privately-owned vehicle market was on the verge of collapse? Electric vehicles, "shared" mobility, and autonomous vehicles were coalescing into a massive sci-fi convergence!
2017: "Self-Driving Cars Could Soon Save the Average Family at Least $5,600 a Year"
https://money.com/self-driving... [money.com]
2018: "In Two Years, There Could Be 10 Million Self-Driving Cars on the Roads"
https://stanfordmag.org/conten... [stanfordmag.org]
2018: "How Autonomous Vehicles Will Transform Ci
Who approved this? (Score:2)
As a resident of SF, I'd love to know how they got approval to do this? I haven't seen any sort of authorization. I'd love to know who not to vote for over this decision.
Haha. "run" (Score:2)
Yeah, they run. While stopped for hours blocking intersections, but they do run.