Highest production VIN in the wild

[vinnie97]

Old hat, my body is ready for 9k!

 

[CarbonFree]

Fact: Tesla’s statement was not weasel words. The statement was very precise.

The actual per-day and per-hour production rates for the last week were not specified, while burst rates, which are are measured in minutes, were specified.

One cannot extrapolate burst rates to any kind of meaningful overall line rate, so there is no legitimate reason to mention them. Further, the burst rates for earlier days were not mentioned, so there is no way to tell if burst rates improved, declined, or remained constant. No were any rework or quality control rates mentioned. The information is useless.

There is only one reason to divulge weekly rates and minute/burst rates, but not divulge the daily and hourly rates in between, and that is to attempt to mislead by omission. Which is a pretty weasly thing to do.

P.S. I'm happy to announce that my own burst rate for authoring messages on TMC has now exceeded 6,000 a week.

 

[Garlan Garner]

I'm sorry, but the highest VIN numbers posting here are not doing Tesla any favors.

There were 3xxx's in December.

 

[EnergyMax]

I'm sorry, but the highest VIN numbers posting here are not doing Tesla any favors.

There were 3xxx's in December.

So max run rate of about 700 Model3/week, and dropping each day we don't see 9xxx.

(edit: average rate since December. We can really only get insight on multi-month average run-rates, as short-term run rates can be obfuscated by VIN jumping.)

 

[ZeApelido]

So, which is your agenda here?

Keep in mind I have a good % of my money invested in Tesla stock and have a first day in store reservation for the Model 3.

I'm annoyed by stupidity and lack of common sense. I'm trying to parse through the extreme Tesla fanboi-dom here and the doom-and-gloomers elsewhere.

These statements about production rate and the confusion that followed, and now the defensiveness of some people just shows me a basic lack of common sense, or more likely that some people are so pro-Tesla they can't think critically. Its annoying.

 

[aubreymcfato]

Remember that the Bloomberg tracker is being updated every few days:
We Set Out to Crack Tesla's Biggest Mystery: How Many Model 3s It's Making

 

[mongo]

Remember that the Bloomberg tracker is being updated every few days:
We Set Out to Crack Tesla's Biggest Mystery: How Many Model 3s It's Making

I admit, the updates and higher world visibility for submission does make it useful/ an improvement on the previous lesser known pages.

 

[trentbridge]

I admit, the updates and higher world visibility for submission does make it useful/ an improvement on the previous lesser known pages.

It needs one last enhancement - the "Total Cars Produced" should update by one every ten minutes showing an additional Model 3 has entered the world. Based on 1052 per week.

 

[CarbonFree]

I'm getting a few dislikes to my prior post, so let me explain my treatment of burst production rates.

Firstly, of course, burst does not equal sustained. I can sprint from the living room to the kitchen at ten feet per second. That does not mean I can sustain 163 miles a day. Similarly, a Model S can reach 155 miles per hour, but not 3720 miles in a single day.

But more importantly, burst does not remotely predict production line speed. Imagine we're living before the age of modern firefighting; we have a fire to put out, and have organized a simple simple bucket brigade to do so. Someone operates a sunken well, filling buckets one at a time. The buckets are passed person-to-person in a line, and the water in the bucket is dumped on the fire once it reaches the conflagration.

How many buckets can we dump on the fire in a day? I can pass a bucket from hand to hand in a burst in about a second. Sustained, let's say two seconds. If the line is just me, I can do two buckets a second. If the line is one hundred people long, how many buckets will we do per second? A tiny fraction of that, because the bucket delivery rate is a function of the slowest person in the line.

Can we just find the slowest person and speed them up, or replace them? No, because the slowest changes from second to second. One person may need to use the bathroom. One person may have slipped. One person dropped a bucket. Someone is changing their gloves. Another is getting tired. Another is distracted. Another has a cramp. One wants a drink. The well crank may need oil. At any point in time, several, if not dozens, of people are simultaneously operating at below "sustained" rate, and *far* below burst rate. Once they resolve their problems, others suffer new problems. It's a game of virtual whack-a-mole.

You can compensate a little bit for this by piling up buckets next to you when the person ahead isn't ready to take a bucket, but that's not a solution, because you can't clear that backlog quickly. Rarely, you'll be able to clear some of the backlog at a burst rate, but that just moves the pile to other points downstream. Most of the time you can't truly remove the backlog from the line it at anything faster than the average rate of the line, which, as we've discovered, is quite slow.

Tesla is quoting the rate at which a single person can move buckets from a pile to another pile. It's not the rate that the production line can move, and it's not even a rate that the particular benchmarked station will ever see in production. Burst is entirely artificial. Even sustained is artificial. The real rate is a function of the rate probability distribution functions of all elements in the line.

 

[mongo]

The actual per-day and per-hour production rates for the last week were not specified, while burst rates, which are are measured in minutes, were specified.

Not in standard manufacturing reporting.

Further, the burst rates for earlier days were not mentioned, so there is no way to tell if burst rates improved, declined, or remained constant.

Trend is a separate issue. In the time quoted, with the volume quoted, the rate, when extrapolated to the time period quote, would produce the number of vehicles quoted.

In the last seven working days of the quarter, we made 793 Model 3's, and 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.

Tesla made 793 vehicles in 7 days, in the last portion of those days, the build rate matched what you would need to make 1,000 vehicle in a week. Any claim that they were taking about about sub hour windows in those days requires corroborating data.

But more importantly, burst does not remotely predict production line speed.

When done properly, it does, and you have given no data to support the claim it was done improperly. Burst production is virtually identical to run at rate tests which are a standard manufacturing line verification technique.

 

[CarbonFree]

Burst production is virtually identical to run at rate tests which are a standard manufacturing line verification technique.

Not according to Tesla's Q3 letter:

Several manufacturing lines, such as drive unit, seat assembly, paint shop and stamping, have demonstrated a manufacturing ability in excess of 1,000 units per week during burst builds of short duration. Other lines, such as battery pack assembly, body shop welding and final vehicle assembly, have demonstrated burst builds of about 500 units per week and are ramping up quickly. Likewise, cell production at Gigafactory 1 continues to ramp, and current cell production makes it one of the largest battery cell manufacturing facilities in the world.

Here, they measured burst independently on different parts of the M3 production line, and the period was not measured in days, but rather “short duration.” Similarly, in the Q4 update to Model 3 production, they mention a rate on "each of our manufacturing lines," but there is only a single line for some portions of M3 assembly, which implies they are still measuring burst on disjoint portions of the process, independently of each other.

 

[mongo]

Not according to Tesla's Q3 letter:



Here, they measured burst independently on different parts of the M3 production line. Similarly, in the Q4 update to Model 3 production, they mention a rate on "each of our manufacturing lines," but there is only a single line for some portions of M3 assembly, which implies they are still measuring burst portions of the process independently.

The Q4 update says all lines ran at 1k/wk burst, an improvement of the Q3 letter of 2 months previous which stated some were only at 500/ wk. Since they specifically state all lines, that means total production can do that.

Independent / non-simultaneous testing is fine as long as the separate test do not have interdependencies. If you can make packs (when you have cells) at 1k/wk and can assemble cars (when you have packs) at 1k/wk, then you can make cars at 1k/wk as long as cell production and transportation keep up. But yah, you can only build as fast as your slowest process, like Elon has mentioned many times.

Q3 -> different run rates, some less than 1k/wk at ~500, and some even worse (GF). Clearly called out.
Q4 -> different run rate, none less than 1k/wk. Also clearly stated (perhaps some ambiguity if it includes non-Tesla suppliers, but there has not been much news to say they are slow (other than potentially the headliner thing))

 

[davedavedave]

TLDR... Amdahl’s law. ;-)

 

[Twiglett]

I'm getting a few dislikes to my prior post, so let me explain my treatment of burst production rates.

Lets not eh
Can't speak for the dislikers personally, but maybe the dislikes are because its a really (really) long post on an off topic subject that they want moved somewhere other than this thread which is about VINs in the wild?

 

[bnsfengineer]

Highest production VIN in the wild???

 

[JRP3]

Yeah arguing about the details of statements from over a month ago probably isn't useful and is off topic.

 

[Sandiegodoug]

My brother picked up 77xx last week in marina del Rey

 

[chrishu]

I think at the moment for tesla the model 3 ramp up, while limiting the cashburn, are the 2 most important tasks. So that will probably mean, checking if they can do a 1000 weekrunrate, for a week or 2, and when that works, go towards getting up to a 2500 week runrate, even if that means , lowering the runrate in the meantime, to less then 1000, because of maintenance on the different lines that needs to happen to get to the higher 2500 runrate ( also if they wanted to keep a higher (1000) runrate in the meantime, would mean a higher cashburn then is probably justifiable ). So I think that quarter 1 deliveries for the model 3, will probably be between 6000 and 8000, and likely closer to 6000 then to 8000. Tesla is also going in uncharted territory as model 3, will be their first model, that will produce at a 2500 runrate, which means more then 2500*13 weeks=30000+ cars a quarter. So probably the most important question on the q1 earnings call, will be how that new grohman battery line is coming along.

 

[acuma50]

VIN 51XX delivered yesterday in Springfield, NJ, within a week of the earliest available delivery date. Midnight silver metallic, long range, 18” wheels. I already love the car, although it’s my wife’s ;-). It handles like a dream. The low profile interior is great and there’s a lot of room to put things (unlike the S). Having the instrument cluster and all the controls on the center screen takes about 5-15 seconds to get used to despite what you see on YouTube ;-) I can already tell that the D version will overall be a better car but I didn’t want to wait and the rear wheel drive is fine (good enough). The wheels slipped a little bit while accelerating aggressively but it was raining out. My S D wheels haven’t slipped even in the rain.

 

[UkrHog]

View attachment 282767 View attachment 282770 VIN 51XX delivered yesterday in Springfield, NJ, within a week of the earliest available delivery date. Midnight silver metallic, long range, 18” wheels. I already love the car, although it’s my wife’s ;-). It handles like a dream. The low profile interior is great and there’s a lot of room to put things (unlike the S). Having the instrument cluster and all the controls on the center screen takes about 5-15 seconds to get used to despite what you see on YouTube ;-) I can already tell that the D version will overall be a better car but I didn’t want to wait and the rear wheel drive is fine (good enough). The wheels slipped a little bit while accelerating aggressively but it was raining out. My S D wheels haven’t slipped even in the rain.

Awesome! Congrats looking forward to getting ours, hopefully my wife will let me drive it around