Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla Q3 2017 Vehicle Deliveries and Production

This site may earn commission on affiliate links.
I don't mean to be a total nerd, but that is not how factorials work. A factorial means you are applying it to the *entire* curve... so saying they are off by a factor of 10 infers that all of their numbers are to be divided by 10. I don't think this is what you mean, is it?

88E8B59D-40E1-4705-8379-5C6B80D31123.jpeg
 
  • Funny
Reactions: dsvick and jhs_7645
I don't mean to be a total nerd, but that is not how factorials work. A factorial means you are applying it to the *entire* curve... so saying they are off by a factor of 10 infers that all of their numbers are to be divided by 10. I don't think this is what you mean, is it?
Well, if you want nerdy...

The thing is that the growth curve is exponential at the beginning and since the derivative of an exponential function is still an exponential function. That means that being X weeks late also means growth lower by x weeks. So as long as both graphs, projected and real are in their exponential phase the factor by which the real production graph is lower should not change too much.

model-3-s-curve.jpg


Sorry for the very cheaply edited photo, it should show a 4 week delay and the consequences on deliveries. Look at the end of September and the end of October. See that the factor between those two doesn't really change too much? That of course changes once the growth stops being exponential and even by the end of October it should be lower than ten, but it will still be quite big. By the end of November it should only be two and in late January it should be the same.

But if they were real exponential graphs the factor should always be the same. Simple example 10^n is always ten times more than 10^n-1, the same is of course true for 1.3895^n and 1.3895^n-7. And if n would be weeks and 1.3895 the growth factor per week, a seven week delay would be the same as always missing projected deliveries by a factor of ten.

Now since the S curve is only nearly exponential to a certain point, Tesla has the chance to catch up.
 
Last edited:
  • Like
Reactions: Waiting4M3
Well, if you want nerdy...

The thing is that the growth curve is exponential at the beginning and since the derivative of an exponential function is still an exponential function. That means that being X weeks late also means growth lower by x weeks. So as long as both graphs, projected and real are in their exponential phase the factor by which the real production graph is lower should not change too much.

View attachment 251596

Sorry for the very cheaply edited photo, it should show a 4 week delay and the consequences on deliveries. Look at the end of September and the end of October. See that the factor between those two doesn't really change too much? That of course changes once the growth stops being exponential and even by the end of October it should be lower than ten, but it will still be quite big. By the end of November it should only be two and in late January it should be the same.

But if they were real exponential graphs the factor should always be the same. Simple example 10^n is always ten times more than 10^n-1, the same is of course true for 1.3895^n and 1.3895^n-7. And if n would be weeks and 1.3895 the growth factor per week, a seven week delay would be the same as always missing projected deliveries by a factor of ten.

Now since the S curve is only nearly exponential to a certain point, Tesla has the chance to catch up.

So.... Not factorial.
 
I both agree and disagree with your statement. When BMW says deliveries of the new 3 series have commenced there are not a trickle of them coming out of the factory. They don't make that statement until they are doing volume production and they are never late, at least not in the last several generations of vehicles.

Tesla might be having supplier problems or component failure problems or maybe just problem with final assembly. I agree that they have a lot of experienced people involved but they also seem to have more than their share of problems for an auto maker with so many auto industry all stars on the payroll.

So yes, you're correct. But there are a couple of differences in the situation. First the series 3 comes off a retooled line with an existing workforce. They're not trying to build up the workforce, construct a whole new line, and prove out the processes for making the new model simultaneously. Tesla is.

Second, and this is Tesla's own choice, they decided not to run pre-production. Or they decided not to call it pre-production. They're effectively running pre-production right now, but claiming it's production. BMW will have spent months proving out the processes before declaring that they're building production car. Even so, don't imagine that BWM doesn't have a LOT of troubles in the first few months of running. It takes them a while to get their lines up to speed. And it's quite common for the 2017 model to become the 2018 model because of various issues. But Tesla makes their promises public. The other automakers don't and so we don't see it.

Why did Telsa do it this way? What would be better, I suppose? Would it be better for Tesla to quietly run pre-production for 4-5 months more and then start shipping in larger quantities?

I contend that Elon does this deliberately, fully knowing the consequences. He does it as a means to motivate the team to meet near impossible schedules. It's one thing to blow a quiet internal deadline. But when a very public guy like Elon goes out there and public announces that XXX will be shipping on XX/XX/20XX, there is no ambiguity or forgiveness. Yes there's much media thrashing and a bunch of folks in this echo chamber get excited when the deadline is missed. But at the end of the day things get done a lot faster than the otherwise would.

Remember that Tesla is still skirting the edge of financial viability. And getting product on the road can be life or death. And in a year, who will care that the 3 was 2-3 months late? Sometimes I don't know if this strategy is deliberate and genius, or just insane. But I'm pretty convinced that it's the right way and the only way. There's not been a new automaker in a century. Without this crazy, nuthouse process, I don't think Tesla would make it either.
 
No. Tesla's own numbers are obviously more reliable than insideevs.com guesses. Here is what Tesla reported to shareholders regarding its first three months of Model S deliveries:

"In Q3, we produced almost 350 and delivered over 250 Model S sedans to our customers, all while maintaining very high build quality levels."

Technically, Tesla did deliver a handful of cars 8 days before the quarter began, so we don't have data on EXACTLY the first three months' sales and production. But these numbers are not more than 8 days off.

Bottom line: At this early stage, the Model 3 ramp is NOT any faster than the Model S and Model X ramps. It is in fact eerily similar to the prior two ramps, despite being designed for ease of production. It is early, we can hope for faster ramping in future periods. But it's not happening so far.

Here's a link to Tesla's report:
Tesla Motors, Inc. Third Quarter 2012 Shareholder Letter
I agree with the above but 250 delivered for the third quarter, even allowing them June when deliveries actually started is still far less than the 300/month you initially quoted. I have owned and been investing with Tesla since 2010. What they are doing is both very hard and very impressive. But they have missed guidance and production ramp now in 4 out of 4 vehicles. I do not fault them as I believe they are doing the impossible and thus cut them slack. I fully believe they will work through the problems and issues. But to say they are not and have not been late in the past is trying to whitewash history. I prefer us to stick to the facts.
 
I don't mean to be a total nerd, but that is not how factorials work. A factorial means you are applying it to the *entire* curve... so saying they are off by a factor of 10 infers that all of their numbers are to be divided by 10. I don't think this is what you mean, is it?

Don't mean to be a grammar geek either, but it would not "infer". It would "imply" and you may "infer" the meaning.,
 
  • Like
Reactions: R.S and jhs_7645
You are correct..

So what did you actually mean? That we should say 4 weeks late because the growth isn't really exponential and we need to look more long term? Or just because 4 weeks late doesn't sound so bad? IMO it doesn't really make a difference, but if you prefer saying they are 4 weeks late...

I doubt Tesla will deliver many Model 3s next month, because they are 4 weeks late. And while I am not 100% certain that delay won't increase, I'm relatively positive that they will deliver more than 1000 cars in October.
 
So what did you actually mean? That we should say 4 weeks late because the growth isn't really exponential and we need to look more long term? Or just because 4 weeks late doesn't sound so bad? IMO it doesn't really make a difference, but if you prefer saying they are 4 weeks late...

I doubt Tesla will deliver many Model 3s next month, because they are 4 weeks late. And while I am not 100% certain that delay won't increase, I'm relatively positive that they will deliver more than 1000 cars in October.

Yes, that is basically what I meant. and yes.. it makes me feel better :). If you go back to my original point, what I meant was that I don't think it's correct to state that they are behind by a factor, nor do I believe that it is correct to say that they are 1200 cars behind, a better way to describe where they are at is 'N' weeks behind... that would mean that the ramp is probably still going to be virtually the same, just later (time shifted) than originally planned. I confused the topic by mis-using the word 'factorial' but my point remains the same.

I think this horse has been sufficiently beaten though...
 
I know everyone is anxious; I know how hard it is to wait for something we look forward to ....

It appears some, however .... do not endure the wait well.

Would't it be for better for Tesla to "get it right?"
I cannot imagine the complexities of producing a brand new car in a brand new factory with employees that have never built this model previously (and some are even new to auto manufacturing).

I am completely confident that once the production sequences are ironed-out; the robots fine tuned; the inspections and quality control procedures finalized, Tesla will be able to accelerate production and produce a truly fine, problem-free automobile.
Let's give them the opportunity to make it right from the get-go.
 
  • Like
Reactions: insaneoctane