Daniel in SD
(supervised)
True.
When do we get to see the 2019 CA disengagement reports I wonder? I'm very curious to see if Tesla reported any testing in CA. It seems like they should have...
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Elon: "Feature complete for full self driving this year"
- Thread starter dckiwi
- Start date
When do we get to see the 2019 CA disengagement reports I wonder? I'm very curious to see if Tesla reported any testing in CA. It seems like they should have...
I think it usually comes out mid Feb. And, yes it will be the first report since 2016 where Tesla has more than 0 miles.
Cheburashka
Active Member
diplomat33
Average guy who loves autonomous vehicles
Do we have actual confirmation that Tesla will report autonomous miles or are we just assuming they will because we are assuming they tested "feature complete"? I ask because in the past, Tesla has cleverly avoided reporting any autonomous miles by claiming that they did not actually test an autonomous system on public roads. Tesla might do that again to avoid reporting any autonomous miles.
Don't get me wrong. I hope Tesla reports autonomous miles because it would be great to get some data on the reliability of the FSD that Tesla has in development. It would be nice to be able to attach some numbers to "feature complete". I am just skeptical that Tesla will report any numbers yet.
Daniel in SD
(supervised)
You have to email a request for AV disengagement data now.
They do have AV accident reports, and Tesla is doing EXCELLENT! No crashes in 2019. The GM Cruise had 60! I didn't count Waymo.
They do have AV accident reports, and Tesla is doing EXCELLENT! No crashes in 2019. The GM Cruise had 60! I didn't count Waymo.
diplomat33
Average guy who loves autonomous vehicles
Maybe Tesla will report it. But Tesla might still try to claim that the demo was not actually autonomous and so it does not count like they did before.
The scary part about California AV accident reporting is that could crush the AV technology advancements if California puts in contact information. Most the companies reporting are deep pockets, hence easy money regardless of fault.
EDIT - I looked over a few reports, and there is no contact information for the non AV vehicle. I scanned 10 at random, and no At Fault AV accidents either. Typical accident is vehicles running into the AV.
EDIT - I looked over a few reports, and there is no contact information for the non AV vehicle. I scanned 10 at random, and no At Fault AV accidents either. Typical accident is vehicles running into the AV.
Last edited:
IIRC, Tesla tested a handful of days with poor reports, and as soon as they had a good video, they stopped immediately. Best I can tell, they have not filed anything with California since.
Daniel in SD
(supervised)
There's no way to disprove this theory about lawsuits except to point out that there are probably over a million vehicles in CA owned and operated by corporations with deep pockets.
I wonder if the Waymo and Cruise vehicles get into more accidents than regular drivers even though they aren't at fault? It seems like that would be unacceptable for any widespread deployment of autonomous vehicles. If I drove around erratically and unpredictably I could probably cause a bunch of accidents that weren't technically my fault.
And there are law firms who actively seek out the million business owned vehicles. That's why your business auto insurance is double your home policy.
It's already happened to AV's in at least one case. A light motorcycle ran into the side of a Cruise Bolt a couple years ago, and the police ruled the bike operator at fault. The suit was filed based on the idea that an AV doesn't behave like a San Francisco driver, that is, driving like blind drunk teenager who's in a hurry. The AV Bolt was obeying the law, the MC was not. I will admit to not knowing the outcome, but 'somehow' it made it to the national media outlets. This is usually because a legal firm sent out press releases. A minor traffic accident is not national news.
OPPSS! International news, not national. Here is one of hundreds of articles about it: GM sued over self-driving car crash with motorcyclist
Last edited:
I don’t think tesla has that big of a head start on full self driving as we think. I seen kia sonata has summon now.... and quite a few cars offer steering assist with adaptive crews control.
Daniel in SD
(supervised)
GM settled: GM settles lawsuit with motorcyclist hit by self-driving car
It does sound like an interesting case. I guess I'd have to see the video but I feel like if I was on the jury I would put more blame on Cruise than on the motorcyclist. I guess the question I'd ask myself is whether or not the accident would have occurred with the average human driver driving. In California you are required to avoid accidents if you can.
So what do you think would happen if the 100+ accident reports in 2019 in San Francisco by AV's included contact information like the original reports did? You don't believe there is anybody in California would see that as an opportunity to gain wealth easily?
Daniel in SD
(supervised)
I think ambulance chasers are already pretty efficient and people are plenty litigious in California already. I'm skeptical it would make any difference at all.
I dunno, kinda hard to claim it wasn't autonomous testing or a demo when it was shown during an event called "autonomy day".
Still can't even read speed limit signs...but we have trashcans (well some do)!
Cheburashka
Active Member
I read Tesla's statement about this a few years ago. I'm not sure they do this anymore, but maybe.
From a programmer/developer perspective, my immediate thought is that this is gotta be the most stupid way to solve the problem I've ever seen. It cannot be a permanent solution.
The problem is that your radar see something that is at a different place (such as overhead sign), and votes the location for whitelisting. But there is a reason they need that, likely because there is no way to distinguish the radar signal from a stationary vehicle. Now they just made the system more fragile by introducing thousands of edge scenarios, sometimes the car sees something not there (phantom braking), or your car ignores something there (stationary vehicle under an overhead sign).
I think this will also be the case with HD maps. You can use them as guidelines. If the road has 3 lanes, but covered with snow, people still drive in these invisible lanes because they've driven there before (human HD map / familiar). The AV could not know this without that map. But you cannot rely on them for safety, or override drivable areas or emergency braking ,because the HD map says its safe. Perception needs to be the main rule.
Similar threads
