Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Transportation

Cruise Robotaxis Return To the Bay Area Nearly One Year After Pedestrian Crash (techcrunch.com) 27

Cruise is returning to the streets of Sunnyvale and Mountain View for the first time since it paused operations in the Bay Area after a robotaxi struck a pedestrian in October 2023. From a report: The company said Thursday that it will put "several" vehicles driven by humans in the two cities that will initially perform mapping. The company said it hopes to progress to supervised AV testing of up to five robotaxis "later this fall."

"Resuming testing in the Bay Area is an important step forward as we continue to work closely with California regulators and local stakeholders," the company said in a post on X. "This will allow our local employees to engage directly with our product as they refine and improve our tech through R&D." The decision to bring Cruise's autonomous Chevy Bolts back to the Bay Area comes just a few months after the company reached a settlement with California's Public Utilities Commission (CPUC). As part of that deal, Cruise paid a $112,500 fine for failing to provide full information about the October 2023 crash.

This discussion has been archived. No new comments can be posted.

Cruise Robotaxis Return To the Bay Area Nearly One Year After Pedestrian Crash

Comments Filter:
  • by Joe_Dragon ( 2206452 ) on Thursday September 19, 2024 @03:55PM (#64800859)

    perform mapping? so each area needs to be super mapped?
    and how often does it need to be updated?
    And what they super map on road and then the next day work starts to change that road?

    • Mapping is how most of the self-driving cars work. It's an incredibly detailed map, including details like the height of the curb. Tesla is the notable exception that doesn't use such mapping (but their self driving tech doesn't seem to work as well either).
      • Tesla is the notable exception that doesn't use such mapping (but their self driving tech doesn't seem to work as well either).

        Accidents per million miles:

        2.8 Human
        1.0 Cruise
        0.4 Waymo
        0.13 Tesla

        • Given the way these companies bitch and moan and argue that everyone but themselves caused all the accidents they actually were involved with, I'm not surprised that your stats end up looking fairly reasonable afterwards.
        • That's pretty nonsensical considering a human has to be at the wheel at all times with Tesla.
          • by zlives ( 2009072 )

            ouch, nice catch
            i would also like to understand the scope of the human data, is it only the same roads and driving conditions that the "self" driving companies do their miles in. how many of those miles are in the parking lot or track.

  • From what I hear (Score:3, Interesting)

    by rsilvergun ( 571051 ) on Thursday September 19, 2024 @03:59PM (#64800871)
    They are all over Phoenix Arizona now and they work perfectly well. At this point I think we are just waiting to see how long it takes to work out the legal liability issues around self-driving cars. I don't think that'll take more than 3 to 5 years.

    We've got around 7 million professional drivers of the light vehicle kind, many of who do it because there are otherwise unemployable.

    I'd say within the next 5 years, 10 of the most every single one of them is going to be completely and totally unemployable. Even if we want to pretend there's magically going to be 7 million new jobs for them nobody's going to want to hire them for those jobs any more than they want to hire them for a job today.

    We need to start thinking about what we're going to do with 7 million completely unemployable people because if we don't somebody else will and they're going to do bad things to us with them...
    • You could say the same thing about the longshoremen who loaded & unloaded ships before containerization. They found jobs, shipping got faster and much cheaper, and fewer product is "lost" either through theft or ineptitude. No I'm not worried about huge proportions of society being unemployed when something better comes along.
      • Sure I will (Score:4, Interesting)

        by rsilvergun ( 571051 ) on Thursday September 19, 2024 @04:35PM (#64800973)
        Here's an article on the effects of 50 years of non stop automation: https://www.businessinsider.co... [businessinsider.com]

        The difference here is that was 50 years of slow, steady automation. We at least found shitty service sector jobs for most of them and made up reasons why they rest were disabled. Settling in to permanent 8-10% unemployment and cooking the books to make it look like half that.

        We're gonna double that in 10 years instead of 50. While also replacing ton of folks with AI. And with no new jobs on the horizon to replace them.

        If we don't do something about it we'll have the same sort of social unrest every country with 15-20% unemployment has. We know this. The only question is what, if anything, we're gonna do about it.
    • by CEC-P ( 10248912 )
      Yeah, deserts are easy mode. Try and this in snow let alone rain. Humans don't even know where the lines in the road are for most of the winter.
    • by gweihir ( 88907 )

      I agree. Not 10/10 of drivers, there will be specialty jobs and people that pay extra for a human driver. But 9/10 will be out of a job because self-driving cars just perform better and cause less damage. Incidentally, this is not unexpected. Eventually, human drivers will just look like a menace in comparison to self-driving vehicles. They already are in many scenarios, but people are still used to human drivers killing people while they are not used to self-driving vehicles killing far fewer people (by an

    • by sixoh1 ( 996418 )

      I've tried reaching out to local and state politicos and NFIB and other 'small business' interest groups about this - automation would tend to move in the direction of capital because it costs money to purchase a robot or robo-taxi etc. - but the concentration of capital that results from this is toxic in the long run for the community and the businesses that use automation. At some point not only will the low skilled labor jobs disappear, but the next ripple is that the technicians to keep those very self-

    • At this point I think we are just waiting to see how long it takes to work out the legal liability issues around self-driving cars.

      I don't think there's any question to be resolved there. I know Waymo has stated that they accept all liability, just as a human driving the vehicle would, and I believe others have agreed to that as well. And, really, there's no other possible allocation of liability, at least for level 4/5 systems where the system is fully in control with no human backup.

  • A human driver kills another person on a weekly basis in the bay area, yet we still allow humans to drive. No suspension, no forced retraining of humans, nothing. Correct thing to do is every time a human kills someone, all driver licenses must be suspended as punishment and everyone should have to go through training before being allowed on the road.

  • I'm taking pre-orders for my new anti-dragging, kevlar lined, chainmail clothing line.

    • by zlives ( 2009072 )

      some believe in defense is the best offense, for others, let me present you my directional emp canon. I call it AutoPilot, because that makes sense.

      • The chainmail offers EMP protection. You definitely want to be wearing some when you fire that thing, keep your phone in the inside pocket. Street fighters that survive wear chainmail.

  • At least for myself, the Cruise app is still "offline" in SF.
  • These will never be 100% safe.
    • Neither will humans. What's your point?

      • So you prefer having your agency taken away from you? Just get into the potential death machine with no way to control it yourself and hope that it doesn't fuck up and get you killed?
        Worse: you get to sit there in utter horror as it mows down a pedestrian, when you could have easily stopped a normal vehicle yourself and avoided killing an innocent person?
        Do you want to risk your kids to one of these machines? Don't tell me that's an 'eppeal to emotion' fallacy because it's 100% applicable.

        You either don
        • If the AI cars have a better safety record than humans by a large factor, what then? Shall we continue to give people the agency to run each other down because....why? Or how about we give them the agency to choose to be driven around by a safer AI driver?
          While currently the AI cars only perform well in certain cities/weather conditions, they do perform MUCH better than human drivers in those conditions. Look at the stats. Now imagine how much better the tech might be in a decade or two.
          Demanding 100% perfe

Experiments must be reproducible; they should all fail in the same way.

Working...