Document
Keyword: "Elon’s tweet does not match engineering reality per CJ."
Keyword: "Elon’s tweet does not match engineering reality per CJ."
Last edited:
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
"Elon’s tweet does not match engineering reality per CJ." - CJ Moore, Tesla's Director of Autopilot Software
- Thread starter Bladerskb
- Start date
powertoold
Active Member
Interesting, can anyone explain why that portion, "does not match engineering reality" is whited-out / blanked-out / redacted in the actual document?
Ssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssshhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh...ehhSsssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssshhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhocking.
Wudyamean l5 ain’t happening next week?
Wudyamean l5 ain’t happening next week?
As an Engineer I definitely plan on using the line "does not match engineering reality" when dealing with people from marketing who overstate things, and I don't even work for Tesla.
DJVoorhees
Member
Document
Keyword: "Elon’s tweet does not match engineering reality per CJ."
Even they know Elon is a huckster. I feel for the folks who paid thousands of dollars for summon and broken auto park.
diplomat33
Average guy who loves autonomous vehicles
"Driver interaction" seems like a much broader term than disengagement or even intervention. I don't think it would include applying torque to the wheel as that is a driver monitoring requirement. IMO, "driver interaction" probably includes stalk confirmations, corrective braking, accelerating and steering and of course disengagements.
The 1M or 2M miles per driver interaction appears to be a rough ballpark number that Tesla is aiming for in order to reach L5. I don't think it is possible to actually measure or classify levels by their disengagement rate. I know the SAE document says that you cannot quantitatively measure the SAE levels.
In any case, the current FSD Beta is nowhere near 1M or 2M miles per driver interaction. So Tesla is very far from their own stated L5 goal. But it is sneaky that Tesla won't release any data on this. That way, there is no data to quantify how far Tesla is from their L5 goal and they can just ride on the hype instead.
I wish Tesla would release the current miles per interaction so that we could get a sense of how far away Tesla is from their L5 goal. Tesla could release the data of FSD Beta miles and number of driver interventions every quarter and we could see how much progress Tesla is making.
Last edited:
What do you mean when you say “their L5” goal? Does Tesla have a different set of goalpost for L5 than the SAE? If so shouldn’t Tesla not say L5 but rather call it something else to avoid confusion?
EVNow
Well-Known Member
"Senior Software Engineer" is surprised engineers think comments by Execs don't match engineering reality. I guess you have never worked in a tech company. I can tell you in my 30 years as a "senior software engineer" - comments by execs has *never* matched engineering reality.
diplomat33
Average guy who loves autonomous vehicles
No. I think we can assume that Tesla is using the same definition of L5 as the SAE is.
Here is what the SAE says about L5:
The issue is that the SAE defines L5 but does not provide specific metrics for actually measuring when you have achieved L5. So how do you determine when you system meets that definition above?
So by "their goal", I simply mean Tesla's way for measuring when they think they will achieve the SAE definition of L5 above. The "1M - 2M miles per driver interaction" appears to be Tesla's metric that they made up to help them determine when they think they will achieve the SAE definition of L5.
diplomat33
Average guy who loves autonomous vehicles
Yes, "City Street Pilot" is FSD Beta. The email is dated March 9. So the 824 vehicles in FSD Beta is from March 9. The email mentions that by the end of the week of March 9, FSD Beta will expand to 1,600. It's now May, so FSD Beta may have expanded to more than that.
Soda Popinski
Member
To be fair, Musk is the Technoking of Tesla, though I don't believe he's the Chief Engineer, like at SpaceX.
There's a big difference between releasing FSD at L2 and FSD at L5.
I think there's a decent chance that Tesla will release FSD City Streets at L2 (requires constant monitoring) to the general public this year. But I doubt it will be good enough to meet the L5 definition.
I think there's a decent chance that Tesla will release FSD City Streets at L2 (requires constant monitoring) to the general public this year. But I doubt it will be good enough to meet the L5 definition.
Absolutely. If it's released to the general public in the form of a download button or otherwise it'll clearly be L2 for the foreseeable future and possibly forever.
Tam
Well-Known Member
Agreed.
I suspect the reason for keeping the driver intervention numbers secrete is because it's nowhere near L3 and above any time soon. That means its rate is just like the current wide release of Autopilot or even worse.
The current revelation of FSD beta from youtube videos shows that FSD beta drivers might have to be more vigilant than plain Autopilot drivers.
But even that makes Elon a liar- he says they are waiting for the march of 9's to release the FSD beta:
"This is a “march of 9’s” trying to get probability of no injury above 99.999999% of miles for city driving. Production Autopilot is already above that for highway driving." CJ's email makes it clear that they are nowhere near that. Elon's number is 1 in 100 million miles, and CJ says they aren't at 1:1M. So there is zero way Tesla can expect to release the beta button "next month" if they need 3 orders of magnitude improvement in reliability before release. It also really calls into question if base highway AP is that high as well.
Last edited:
Yes, but also:
"Tesla said that as of March 9th there were 824 vehicles in the pilot program, including 753 employees and 71 non-employees."
71 non-employees. .007% of all cars. 90% of all the cars running it are employees. It's nothing.
Tam
Well-Known Member
Agreed.
It would be more reasonable for Elon's enthusiasm if hundreds of thousands or millions of Tesla owners have been running the FSD beta a long time ago and we should not be just still in a planning stage of when the download button would still appear someday!
It just doesn't add up that 1 million robotaxis would run around without drivers by 2020 (last year).
Dirty Tesla (Youtube person with FSD beta) is at 3 miles per disengagement.
Only a 1 million time increase in reliability is needed to hit CJ's 1-2 million target. Only 30 million needed to hit Elons 1:100M.
Let's say we manage to increase reliability by 50% every week.
3 to 1 million is only 31 weeks away!
See, we're almost there. Just need to improve exponentially for the next 8 months on something that has taken 5 years to get from 0-3.
Then Elon's number is only 12 weeks behind that!
Ok, Ok, 50% is pretty unrealistic. Let's say we can do 10% per week...
Well, then 1:1M is only 134 weeks away. 2.6 years. Sounds perfect for Elon. It's always 2 years away.