Cruise's disengagement stats

Google
co-CEO

Go to company page Google

co-CEO
Apr 17, 2021 13 Comments

Everyone keeps repeating how disengagement stats don't matter and how Cruise has a crappy tech stack, but if so, how come they manage to have a comparable number of miles per disengagement in SF as Waymo does in the suburbs, per CA 2020 report?

Are they driving the same route all the time? Underreporting disengagements? Training their safety drivers to take more risk?

Asking genuinely, why does the industry seem to disbelieve what looks like Cruise's lead?

comments

Want to comment? LOG IN or SIGN UP
TOP 13 Comments
  • Cruise
    yehxbj

    Go to company page Cruise

    yehxbj
    We did not have a single reportable disengagement in the second half of last year. Beat that.
    Apr 17, 2021 2
    • Intel
      VHarkonnen

      Go to company page Intel

      VHarkonnen
      I didn't either
      Apr 17, 2021
    • Uber
      ewesirnaam

      Go to company page Uber

      ewesirnaam
      Ok, either you’re sitting on a huge gold mine and opting not to use it, or there’s clearly a third variable. Could be running the same carefully chosen tests, could be carefully defining what qualifies as reportable. If the car was perfect to the point where you couldn’t break it in 6 months of honest effort then it would be released to the public, yet it isn’t.
      Apr 19, 2021
  • Google
    paused.

    Go to company page Google

    paused.
    The only thing that matters is for a company to deploy a fully autonomous fleet (no driver) to be consumed by the general public at scale. Everything else is simply noise
    Apr 17, 2021 5
  • Waymo / Eng
    car2quant?

    Go to company page Waymo Eng

    car2quant?
    Without spilling too many beans, the miles per disengage stat isn't the long pole on the way to a commercially viable product. And the tail is very, very long.

    There's probably no way to know for sure who is in the "lead," or whether that's even a meaningful question, but in fairness we have hit a fair number of milestones already that haven't been duplicated yet.
    Apr 17, 2021 2
    • Google
      vbzxg

      Go to company page Google

      vbzxg
      care to elaborate? what do you think is the main blocked for waymo going ipo? is it too early to join a av company (risk on providing viable product) or too late (already too many players in this track)
      Apr 18, 2021
    • Waymo / Eng
      car2quant?

      Go to company page Waymo Eng

      car2quant?
      I assumed this thread was about who is in the lead in terms of technology and capturing future market share. That's what I was responding about. IPO prospects is a bit of an orthogonal question. I would say join Waymo if you're passionate about making robot cars that drive themselves, saving lives on the road, etc. Don't join thinking it's a get rich quick scheme.
      Apr 19, 2021
  • Google
    zczq38

    Go to company page Google

    zczq38
    People misunderstand these stats. They’re “reportable” disengagement and that’s the key word. Basically it’s how often a horrible accident would occur if they didn’t take over.

    Actual disengagement rates for nearly all companies in SF is more like every few miles.
    Apr 18, 2021 0