Tesla competes with every single auto manufacturer. Just because their product takes a different approach, they are ultimately addressing the same customer base and customer need.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Highest production VIN in the wild
- Thread starter insaneoctane
- Start date
-
- Tags
- Model 3
Tesla competes with every single auto manufacturer. Just because their product takes a different approach, they are ultimately addressing the same customer base and customer need.
My question there would be: "How many working days does Tesla have in a normal work week?"
When I read that line originally, I assumed (incorrectly?) that they had a typical 5-day work week. So, 793 cars in 7 days equals about 114 cars per day. So, in a 5-day week, that would mean a run rate of 570/week.
Esme Es Mejor
Member
Google. And the thieving bastards at Uber, who look like they’re imploding, but may pull out of it under new management.
And BYD. And Geely. And possibly SAIC & BAIC, among others.
Being a burst rate shows the line can produce at a certain rate, but does not show the supply base is able to support that. I.e. bunch of parts batched up, run the line quickly, stop when you run out if parts. Saying all lines were at 1k/wk shows internal processing/ production is capable, not just the final line.
If burst production hits your target in actual output(not just rate) that's a really good thing. It means they can hit production numbers in less time than required/ available. That said, they are shooting for 5k-10k per line, so 1k should be burst.
Equivilent number of days per week is a valid point, but there are a couple others to concider as well. The 570 nunber assumes 5 normal vs 7 reported, that the line up time is the same for the reported 7 days as a normal week, and that the build rate was the same each day.
At the end of 2017, they were parts limited (packs, possibly others). So they can only burst build to stock on hand. Because of this, I'd expect they only ran enough parts to test the line to correct issues (as opposed to running flat out on the first trial and then not being able to run more tests till more parts show up). So small bursts in the tens at most.
As they get more sections of the process working, following sections have more parts to work with and can be tested. (Can't finish without packs and bodies, can't have bodies without assembly, the BIW, welding, DU... etc) so there are multiple sub lines that need to operate at the final speed for continuous production. (This can also lead to staggered burst builds which would cut average build rate. )
To me, the last few days build rate comment says there was an increase in rate at the end. 7 days made 793, but the last few days made (1000/7*3) = 429 of those (over half the 7 day production). The all lines working at rate comment says to me that there were no part production/ processing restrictions internal to Fremont (likely also GF) preventing 1k/ week.
Many unknowns, but I don't see anything that contradicts 1k/wk ability (given external parts) We have 0 clarity on how long each line needs to run in a day, but these statements point to less than 24 hours
Lame analogy warning:
A stock gives out quarterly dividends of $1, $1.25, $1.5, $1.75 what would be the expected next dividend? (average, linear extrapolation, max value, other). What of 1.0, 1.1, 1.6, 2, 2.2, 2.3?
JRP3
Hyperactive Member
You're right of course. I was just pointing out that the quoted statement "in the last few days, we hit a production rate on each of our manufacturing lines that extrapolates to over 1,000 Model 3's per week." was not a specific statement of them making 700 cars in the last week.
The bot is from Reddit user blfire: its weeks that is not working well, it never registered >4000 VINs.
I just asked the guy.
Waiting4M3
Active Member
To me the simplest explanation of burst vs sustained rate is the utilization rate of the line. As you described, parts availability could be a constraint, and is an external one, this is both good and bad, good because that means Tesla's automation is working, bad because it's less within Tesla's control.Being a burst rate shows the line can produce at a certain rate, but does not show the supply base is able to support that. I.e. bunch of parts batched up, run the line quickly, stop when you run out if parts. Saying all lines were at 1k/wk shows internal processing/ production is capable, not just the final line.
If burst production hits your target in actual output(not just rate) that's a really good thing. It means they can hit production numbers in less time than required/ available. That said, they are shooting for 5k-10k per line, so 1k should be burst.
Equivilent number of days per week is a valid point, but there are a couple others to concider as well. The 570 nunber assumes 5 normal vs 7 reported, that the line up time is the same for the reported 7 days as a normal week, and that the build rate was the same each day.
At the end of 2017, they were parts limited (packs, possibly others). So they can only burst build to stock on hand. Because of this, I'd expect they only ran enough parts to test the line to correct issues (as opposed to running flat out on the first trial and then not being able to run more tests till more parts show up). So small bursts in the tens at most.
As they get more sections of the process working, following sections have more parts to work with and can be tested. (Can't finish without packs and bodies, can't have bodies without assembly, the BIW, welding, DU... etc) so there are multiple sub lines that need to operate at the final speed for continuous production. (This can also lead to staggered burst builds which would cut average build rate. )
To me, the last few days build rate comment says there was an increase in rate at the end. 7 days made 793, but the last few days made (1000/7*3) = 429 of those (over half the 7 day production). The all lines working at rate comment says to me that there were no part production/ processing restrictions internal to Fremont (likely also GF) preventing 1k/ week.
Many unknowns, but I don't see anything that contradicts 1k/wk ability (given external parts) We have 0 clarity on how long each line needs to run in a day, but these statements point to less than 24 hours
Lame analogy warning:
A stock gives out quarterly dividends of $1, $1.25, $1.5, $1.75 what would be the expected next dividend? (average, linear extrapolation, max value, other). What of 1.0, 1.1, 1.6, 2, 2.2, 2.3?
Other possible constraints on utilization rate are tool maintenance, consumable (nuts/bolts/solder/glue etc) replenishment. It's possible that some machines will need to be calibrated frequently (weekly? or more often depend on the workload), and some consumable replenishment could be a manual process that could require downtime on the order of 1 hour or more. These are things within Tesla's control so they may be able to innvoate beyond industry standard. In factories I've been in (not car industry), people typically shoot for 80% utilization rate. If one wants to be conservative, I think it may be good to factor that 80% in to temper Tesla 1000/wk rate, and call it a sustained 800/wk.
To counterpoint: the 1k rate is only 20% of the planned 5k rate (which, with a 80% target utilization, is really a 6k max rate). So, in this case, 1k may be 1k.
Waiting4M3
Active Member
A month ago a lot of people here would give their left nut for 500/wk, now it's considered "very slow". Some day soon we'll be calling 5k/wk "very slow".
Waiting4M3
Active Member
I would have preferred if Tesla PR could have left out the extrapolated 1k/wk rate all together, and just told us if the 793 in 7 working days is a sustained rate, and that working days means excluding holidays. I would take a rock solid 793 vs a hand-waving 1000.
RTPEV
Active Member
You hit the nail on the head. Tesla intentionally used the extrapolated word so they could specifically work the number 1000 in there. Sounds like many of you don't have a problem with the word extrapolated, but it definitely is marketing-speak murkiness! If a quarterback makes a 60 yard pass in the first minute of a game, you could accurately state that his passing stats extrapolate to 3600 passing yards/game, but please, we all know that quoting an extrapolated stat is not the same as actually hitting that stat. They hit 793 vehicles in the last week, period. Not 1000.
Okay, with that out of the way, I do think they were probably quite close and likely by now are actually at that 1000/week rate.
