Production Rate (incl manufacturing waves)

[timdorr]

In that scenario and for the week ended Sunday, October 21, there were to be 628 cumulative vehicles produced with 385 cumulative customer deliveries. For the week ended October 21, vehicles with VINs ranging between 261 to 566 were delivered to customers. The average is 414 and is likely skewed to a lower, not higher number of vehicles delivered, which makes the cumulative customer delivery estimate of 385 appear likely.

That sounds low to me. Elon said there were 250+ deliveries in Q3, so for there to only be 130 deliveries in the past 3 weeks doesn't really jive with what I've been seeing come through on the spreadsheets. It sounds like they're around 450-500 delivered at this point.

I think your delta of cars delivered and produced is a little high at this point. The 628 seems reasonable for produced numbers (DrComputer say VINs as high as 650 on the line 7 days ago, and the full production cycle is 10-14 days, so it would stand to reason the 650th car is done and in burn-in), but I don't think they have nearly 250 cars in the delivery process. Given how recently small the delivery team was, managing that would have been a nightmare with so few people. The delta should certainly grow in absolute numbers, but on a percentage basis, I think it needs to be smaller or even shrinking over time (especially as they bring delivery ops in-house).

 

[bmek]

That sounds low to me. Elon said there were 250+ deliveries in Q3, so for there to only be 130 deliveries in the past 3 weeks doesn't really jive with what I've been seeing come through on the spreadsheets. It sounds like they're around 450-500 delivered at this point.

My analysis includes information from your spreadsheets. I'm ok with a low number and wouldn't want to risk overstating customer deliveries to date.

I think your delta of cars delivered and produced is a little high at this point. The 628 seems reasonable for produced numbers (DrComputer say VINs as high as 650 on the line 7 days ago, and the full production cycle is 10-14 days, so it would stand to reason the 650th car is done and in burn-in), but I don't think they have nearly 250 cars in the delivery process. Given how recently small the delivery team was, managing that would have been a nightmare with so few people. The delta should certainly grow in absolute numbers, but on a percentage basis, I think it needs to be smaller or even shrinking over time (especially as they bring delivery ops in-house).

Deliveries slowed to a trickle until certain bottlenecks were addressed, including availability of key parts and certain manufacturing deficiencies. There were quite a few vehicles hung up in service (waiting to be completed), and many of these were for cross country delivery. Take a look at posting #1762 and the calculations on how many vehicles can be hung up between completing manufacturing and actual delivery to the customer.

Of the vehicles delivered in California, several VINs were in the 200 range. There haven't been that many over 450 much less over 500.

 

[johngray]

Analyzing VINs vs Reservations - 6 waves of manufacturing?

I did some interesting analysis of Reservation and VIN ordering from Tim's spreadsheet and found some interesting trends.

It looks to me like there have been 6 "waves" of orders fulfilled resulting in a VIN. I have color coded them to show them more easily. And of course, there are always a few outliers - probably due to scheduling deferrals, or other issues. This is a plot of VIN vs Reservation Number for known VINs.

(Sorry for the huge images).

Here is how I interpret this:

• Wave 1 (Blue) was the first batch of SSL's configured. It runs up through VIN 134.
• Wave 2 (Red) starts at VIN 135 but is a bunch of S reservations, but it is scattered through the reservations. It looks like they ran through reservations looking for specific configs, and built those next. Three of them were actually issued during the initial Wave1 but i left them in this wave because they were not SSLs.
• Wave 3 (Green) is the second batch of SSLs. Note the VINs are higher than all the Wave2 S VINs. It looks like after allocating Wave2, they went back and finished all the SSLs.
• Wave 4 (Gray) is a bit random, if you go in VIN order, it just seems to skip around like they had a few configs they could make easily.
• Wave 5 (Blue) looks like they had a few more configs they could make.
• Wave 6 (Orange) looks like cars that were waiting on some part or config that they could not get in volume until recently.

At some point, either Wave 6 or another wave - they will just start generating a nearly straight line as they get all supply.

Here are two other charts - just the VIN numbers in order. The first one is simply SSL vs S. Note they started with S, then Wave2, then one SSL (in the middle of a small gap - wonder if this is when they made the Dealer cars - because none of them would show up in the spreadsheet), then the rest of the S cars made so far.

The second is basically the same view but color coded for Wave (as above) so you can see how they flowed linearly by VIN.

Hope you find this interesting and insightful.

 

[Zzzz...]

Update: my Delivery Specialist emailed me today (unsolicited) to let me know my car is "moving along the assembly line" (no specifics on what stage) and also gave me my VIN, #722. Window 10/25 - 11/8 still stands. Getting closer!

#722 on assembly line is almost perfectly fits bmek chart. Probably even a bit ahead of the "schedule". Would be interesting to know how long car is spending on assembly. Anyone know by any chance how many cars assembled at any given moment?

 

[dsmith2189]

I did some interesting analysis of Reservation and VIN ordering from Tim's spreadsheet and found some interesting trends.

It looks to me like there have been 6 "waves" of orders fulfilled resulting in a VIN. I have color coded them to show them more easily. And of course, there are always a few outliers - probably due to scheduling deferrals, or other issues. This is a plot of VIN vs Reservation Number for known VINs.

Hope you find this interesting and insightful.

I'm looking but can't see the S curve in production that they have always talked about... do you have one that is time based to see it?

 

[bmek]

I'm looking but can't see the S curve in production that they have always talked about... do you have one that is time based to see it?

The graph below is time based (X axis) and you can see the weekly production in the red bars. Production starts slow then accelerates to when it reaches 400 vehicles/week at the end of December. Production is then constant at that rate.

 

[Smallfootprint]

Just adding a data point since we got an update last night: Signature sequence number 385, vin #553 is currently "in assembly where they're getting ready to install the interior"

No estimate was available for how long that will take.

Seems like based on that data point the slope on your blue/cumulative production line might be a touch too high.

 

[Johan]

johngray: your charts and analyzis is spot on! Great work. It must be the way you figure, with them first building most SSLs, then some Ss, then the rest of the SSLs. Would be very interesting to know why. If we had all the characteristics of all the cars delivered by VIN we might be able to spot some pattern here. Probably has to do with supplier issues? The thing is that SSL/S cars don't have that many options that differ. Colour is one, wouldn't delay I'd think? Interior colour - maybe? Performance or not - maybe? Wheels - wouldn't matter I'd think? Almost all I've heard have Pano roof...

 

[ibcs]

To further confirm Signature #121 - Vin #513 is currently in assembly which was held up by the part shortage.

 

[timdorr]

Just adding a data point since we got an update last night: Signature sequence number 385, vin #553 is currently "in assembly where they're getting ready to install the interior"

No estimate was available for how long that will take.

Seems like based on that data point the slope on your blue/cumulative production line might be a touch too high.

They didn't give you a delivery date range at all? That's odd, since you have a VIN at this point.

I'm going to put in 11/7-11/21, since that seems to be the current default date range people are getting.

 

[Smallfootprint]

They didn't give you a delivery date range at all? That's odd, since you have a VIN at this point.

I'm going to put in 11/7-11/21, since that seems to be the current default date range people are getting.

Originally the final configuration paperwork listed the VIN as 5000006573 (????), but the photos we received last night have the vin I provided.

The original delivery window, as communicated on 9/20 was 10/13 through 10/27. No updated delivery window has been provided (thus far).

 

[FlasherZ]

I'm seeing a pattern in that they're just adding a few weeks onto existing windows that were on the spreadsheet. It seems those are just estimated dates and until the factory begins producing the car, Tesla really doesn't know anything. At least that's what I'm figuring out from what little communication I'm seeing. There are still some ones way out of whack (in the other thread, P188 has a new window that's two weeks before S1189's "new window", for example).

 

[digitaltim]

The curve will start looking like an "S" once we have a couple of years worth of production time - this ramp up will be compressed on the timeline.

 

[adurstewitz]

One point of note is that the quantity of cars does not relate to the sequence number once were in general production. I've been told that they will be batching the batteries for the rest of 2012.

 

[joefee]

@bmek /timdorr
What do you think the weekly build rate is now?

 

[v12 to 12v]

I'm seeing a pattern in that they're just adding a few weeks onto existing windows that were on the spreadsheet. It seems those are just estimated dates and until the factory begins producing the car, Tesla really doesn't know anything. At least that's what I'm figuring out from what little communication I'm seeing. There are still some ones way out of whack (in the other thread, P188 has a new window that's two weeks before S1189's "new window", for example).

They added a month to my original delivery window as of yesterday. My first delivery window started last Monday, the 22nd. Then they adjusted it to this coming Monday, the 29th, on last Friday. Now I'm mid Nov to the end of Nov. My Res# is ~50 before yours and my original MVPA was October. Your experience may end up differently though. All my theories have gone awry recently.

I wondered if it is because I ordered a car without a Pano Roof. I'd be just as happy with one that had a pano roof if it made a difference. I was torn on that option.

If I knew that I had this much time a couple months ago I would have changed my wheel option, but I thought they were building my car already and I didn't want to do anything that might create a delay.

 

[sublimaze1]

There are still some ones way out of whack (in the other thread, P188 has a new window that's two weeks before S1189's "new window", for example

To that effect, I will say that I am (as I mentioned in my thread) banking on that being a spreadsheet filler: that is, they had to put something in the cell, so put this guestimated date range. Awry was a good term, and I fully expect my date to change to early to mid december at the earliest.

 

[timdorr]

@bmek /timdorr
What do you think the weekly build rate is now?

So, let's play some math games:

On 10/3 we had 359 cars produced. VINs are now being produced in order and 722 was the highest in production 2 days ago and we've seen as high as the 800's reported from factory tours. A safe bet would put them at 750 produced at the end of the week. That's roughly 400 cars in three and a half weeks.

However, there was also a 1-2 week production delay in there, so it's more along the lines of 2-2.5 weeks for 400 cars. That means we're around 160-200 cars a week, which fits in with previous estimates I had made in the past. I'm still confident we can get most of the Sigs finished by the end of next week (which is 11/4). Again, we should see some production models rolling on to the line next week at some point (assuming they're not going to pause a bit while switching).

Note this doesn't include any time spent on inspection and burn-in. According to my definition, a car is "produced" when it comes off the line.

 

[aviators99]

So, let's play some math games:

On 10/3 we had 359 cars produced. VINs are now being produced in order and 722 was the highest in production 2 days ago and we've seen as high as the 800's reported from factory tours. A safe bet would put them at 750 produced at the end of the week. That's roughly 400 cars in three and a half weeks.

However, there was also a 1-2 week production delay in there, so it's more along the lines of 2-2.5 weeks for 400 cars. That means we're around 160-200 cars a week, which fits in with previous estimates I had made in the past. I'm still confident we can get most of the Sigs finished by the end of next week (which is 11/4). Again, we should see some production models rolling on to the line next week at some point (assuming they're not going to pause a bit while switching).

Note this doesn't include any time spent on inspection and burn-in. According to my definition, a car is "produced" when it comes off the line.

I certainly hope there's no "pause". The reason I think there really won't be is because of the 85kWh/suspension batching. There will be very little difference between the Sig and the rest of the productions this year. Also, the fact that we saw that gray marketing car get produced recently shows that they are already have non-Sigs coming off the line. Aside from the badging and paint color, is there anything that my car will have that is not a signature option that has already been produced? I think I remember that the carbon fiber is optional?

 

[NigelM]

I think I remember that the carbon fiber is optional?

CF is a "performance" option only.