News for hackers. Stuff that matters.
Re:How Many Sensors for Autonomous Driving? by ajross on Wednesday June 07, 2023 @01:45PM (#36230400)
FWIW: the numbers on that site show a vehicle with an accident rate that is vastly safer than average. The site is sort of a joke. They launched it with the intent of proving a prior, still lead with that hypothesis as an assumption in their text, and... they couldn't find the goods!

In fact it's so bad that you just have to assume their data is wildly incomplete because the not at fault fatality rate for a car just being in traffic (i.e. Tesla drivers killed because someone else hits them) isn't correctly represented.

Re:How Many Sensors for Autonomous Driving? by autoexec on Wednesday June 07, 2023 @02:03PM (#36230757)
I don't doubt that someone behind "tesladeaths.com" has an agenda, but whatever their reasons for collecting this data, there are a ton of accidents involving Teslas which makes "they don't crash" entirely false.

Yes, some of those links involve teslas being hit by others, and many likely involve cases where the driver is totally at fault. Most of the links are just reports of crashes still under investigation.(this one for example can't even provide the time when the accident took place https://web.archive.org/web/20230304195927/https://www.bigco...)

Still, that's okay, because this isn't a scientific paper, it's just a website collecting data on accidents involving teslas. Each one would still have to be investigated to find out what the cause of the crash was before you could draw much from the raw numbers, but what we can tell unambiguously is that these vehicles do in fact crash. Here's a case where autopilot killed someone just a few months ago (https://abc7news.com/tesla-autopilot-crash-driver-assist-cra...)