Just some questions after reading the article, wonder if anyone with more experience in the industry can provide some insight How often do Brake and Roll tests uncover errors? How long does a BR...
Just some questions after reading the article, wonder if anyone with more experience in the industry can provide some insight
How often do Brake and Roll tests uncover errors?
How long does a BR test actually take in relation to other steps of the assembly process (because if the narrative is this is being done as a time-saving measure it would have to be a viable bottleneck for that to be true)
When pressed on whether Musk himself gave the order to remove the brake-and-roll test, Arnold said, "I don't have anything further beyond the statement."
So the headline definitively states that Musk specifically ordered this to happen, and this is seriously what they're basing it off of? Is there other proof somewhere they didn't report on or I missed, or is this the only thing they're going off of because that seems kind of flimsy.
So I'm an automotive engineer, but I don't work in manufacturing. I can ask one of my coworkers who is ex-manufacturing next week. Our lab is currently on a skeleton crew because of the holiday....
So I'm an automotive engineer, but I don't work in manufacturing. I can ask one of my coworkers who is ex-manufacturing next week. Our lab is currently on a skeleton crew because of the holiday.
Anyway, I don't know anything about this specific test, but in general 'End of Line' or EOL tests can be used to uncover and trace defects. You may not know how to correctly identify issues, but having a log of the data can help establish patterns and identify suspect parts/vehicles after the fact. It's not just about defects in the moment. Suspect parts can only become a problem months or years down the line.
I'm going to provide a bit of an example, but we're delving into the area of THINGS PAPANACHOS CAN'T TALK ABOUT so I'm going to stay super vague. One of the prototype parts I was working on had some manufacturing defects. The nature of the problem meant they weren't detected by the tests that were put in place, and they were nearly impossible to identify once they were installed in the larger system we were testing. Fortunately we had the logs from a different test earlier in their production that had anomalous, but not obviously problematic, readings to identify the parts in question. So we were able to fix our prototypes and resume STUFF.
Basically these tests aren't just pass/fail and then toss the data. They can help characterize the parts and helps us understand issues that we aren't yet aware of.
THAT BEING SAID, I know nothing about this specific test. From the name it seems important, but that's not exactly an informed opinion.
But this
"To be extremely clear, we drive every Model 3 on our test track to verify braking, torque, squeal, and rattle," Arnold said in a statement. "There are no exceptions."
is bullshit. That's only going to catch the most egregious errors. How driving 'feels' is not calibrated and doesn't provide any data for future analysis.
I see a few possible explanations for why they would do this. Obviously I can't read their minds and I don't work for Tesla, so I don't have inside info:
Everything is on the level, this test was unnecessary and just wasted time and slowed down their production.
The test itself was producing a lot of false negatives (failed the test, but wasn't actually bad) without producing any meaningful data
Some other perfectly legitimate reason
The test was taking too long. Damn the consequences.
The test was working inconveniently well. Which is to say, it was detecting defects that Tesla and Musk felt were better ignored. This is the extremely cynical stance.
Some other less-that-legitimate reason
Some combination of the above
As far as how to verify this. No fucking clue. I'll leave that to the author.
Edit: There, I think I fixed all the formatting issues
Edit 2: It's Monday morning, I spoke with my ex-manufacturing coworker and he wasn't familiar with that particular test. He didn't work on brakes.
Just some questions after reading the article, wonder if anyone with more experience in the industry can provide some insight
How often do Brake and Roll tests uncover errors?
How long does a BR test actually take in relation to other steps of the assembly process (because if the narrative is this is being done as a time-saving measure it would have to be a viable bottleneck for that to be true)
So I'm an automotive engineer, but I don't work in manufacturing. I can ask one of my coworkers who is ex-manufacturing next week. Our lab is currently on a skeleton crew because of the holiday.
Anyway, I don't know anything about this specific test, but in general 'End of Line' or EOL tests can be used to uncover and trace defects. You may not know how to correctly identify issues, but having a log of the data can help establish patterns and identify suspect parts/vehicles after the fact. It's not just about defects in the moment. Suspect parts can only become a problem months or years down the line.
I'm going to provide a bit of an example, but we're delving into the area of THINGS PAPANACHOS CAN'T TALK ABOUT so I'm going to stay super vague. One of the prototype parts I was working on had some manufacturing defects. The nature of the problem meant they weren't detected by the tests that were put in place, and they were nearly impossible to identify once they were installed in the larger system we were testing. Fortunately we had the logs from a different test earlier in their production that had anomalous, but not obviously problematic, readings to identify the parts in question. So we were able to fix our prototypes and resume STUFF.
Basically these tests aren't just pass/fail and then toss the data. They can help characterize the parts and helps us understand issues that we aren't yet aware of.
THAT BEING SAID, I know nothing about this specific test. From the name it seems important, but that's not exactly an informed opinion.
But this
is bullshit. That's only going to catch the most egregious errors. How driving 'feels' is not calibrated and doesn't provide any data for future analysis.
I see a few possible explanations for why they would do this. Obviously I can't read their minds and I don't work for Tesla, so I don't have inside info:
As far as how to verify this. No fucking clue. I'll leave that to the author.
Edit: There, I think I fixed all the formatting issues
Edit 2: It's Monday morning, I spoke with my ex-manufacturing coworker and he wasn't familiar with that particular test. He didn't work on brakes.
This seems to be a pretty big deal. Tesla has automated cars, and those automated cars will malfunction horribly if the brake doesn't work!
I fled from reddit in hope to escape from the Musk spam but it already got there too... at least it's negative this time.
Marvelous !