Model III - Your Case: On Time or On Jonas Time?

[dennis]

I'm interested to see what these "special robots" will do that everyone keeps talking about. There could be some advancement but it does sound a lot more like talk than substantive actuality.

Most of the short repetitive tasks that are currently done by humans on the final assembly line. As a simple example, attaching the underbody tray pieces. This is only cost effective if the robots can do the tasks faster than than the humans thereby increasing throughput. In this example, that would mean screwing in the 10-20 fasteners all at once rather than one at a time.

And there is a moderate chance that 2017 (based on your numbers) is still not going to beat 90,000 a year S+X+3 if they slow-build the first round of Model 3.

It looks like you are forecasting a drop in demand for S+X in 2017 since Tesla's current demand based on Q3 results and Q4 forecast is 100K per year. But at least it is an increase from the 62K you forecast for this year.

 

[NicoV]

There's now way the first 115,000 of us that reserved a car before the unveiling will get their car in 2018 2Q or later. They won't let that happen. That would be quite a slow ramp up.....

On the Model 3 page, delivery for new reservations is still expected for mid 18. That would mean the 400k current reservations will be produced by mid 18.

 

[aronth5]

On the Model 3 page, delivery for new reservations is still expected for mid 18. That would mean the 400k current reservations will be produced by mid 18.

Tesla's website now has "Delivery estimate for new reservations is mid 2018 or later" so as much as I'd like to see current reservations filled by mid 18 the wording "or later" can mean just about anything.

 

[NicoV]

Tesla's website now has "Delivery estimate for new reservations is mid 2018 or later" so as much as I'd like to see current reservations filled by mid 18 the wording "or later" can mean just about anything.

'Or later' implies that there is still a small possibility of mid 2018.

 

[dc_h]

'Or later' implies that there is still a small possibility of mid 2018.

 

[ggies07]

Hahaha. This is perfect.

 

[bonaire]

'Or later' implies that there is still a small possibility of mid 2018.

Timeline late 2012 - Tesla Delays Model X Production To "Late" 2014 - Inside EVs

 

[dennis]

Timeline late 2012 - Tesla Delays Model X Production To "Late" 2014 - Inside EVs

And of course we all know how relevant the Model X timeline is to the Model 3.

 

[JRP3]

And of course we all know how relevant the Model X timeline is to the Model 3.

Seems to be quite relevant among the bear crowd, they have to hang their hopes on something.

 

[neroden]

It has to be far more than a few in 17Q3, and then thousands in 17Q4. Financing companies will want to see serious throughput of Model 3 in late 2017 "as promised" otherwise, gaining the funds to buy enough production parts and hire enough workers and build out sales/service centers won't be as easy.

Well, that's a different argument, isn't it? I guess we'd better hope Tesla gets their robots debugged before the end of 17Q3.

My argument is simply that the robots basically go at zero or at full speed. Off or on. Therefore we see either very low production (if the robots simply aren't working right yet and the line has to be stopped repeatedly to tweak them) or very high production (if they are working right).

This seems to be a technological fact. There is no reason to expect the classic human-labor rampup curve. Anyone doing so is doing the worst sort of reasoning by (bad) analogy.

 

[dc_h]

Well, that's a different argument, isn't it? I guess we'd better hope Tesla gets their robots debugged before the end of 17Q3.

My argument is simply that the robots basically go at zero or at full speed. Off or on. Therefore we see either very low production (if the robots simply aren't working right yet and the line has to be stopped repeatedly to tweak them) or very high production (if they are working right).

This seems to be a technological fact. There is no reason to expect the classic human-labor rampup curve. Anyone doing so is doing the worst sort of reasoning by (bad) analogy.

Robots and line will likely need tuning. It will be different challenges, but similar pattern as other rollouts. I think the S-curve will be steeper and longer than the S or X, but it will not be an on-off switch to 500,000 cars.

Just my moderately educated opinion.

 

[dennis]

Well, that's a different argument, isn't it? I guess we'd better hope Tesla gets their robots debugged before the end of 17Q3.

My argument is simply that the robots basically go at zero or at full speed. Off or on. Therefore we see either very low production (if the robots simply aren't working right yet and the line has to be stopped repeatedly to tweak them) or very high production (if they are working right).

This seems to be a technological fact. There is no reason to expect the classic human-labor rampup curve. Anyone doing so is doing the worst sort of reasoning by (bad) analogy.

I think it is highly unlikely that the Model 3 final assembly line in 3Q17 will be mostly robots. Remember Elon referred to this as AD 0.5, and said that AD 1.0 would come a year later. While I do expect more automation than on S/X final assembly line, there will still be many humans working on the Model 3 line in 2017.

 

[3Victoria]

I think AD 1.0 will come as a shock to many.

 

[dennis]

I think AD 1.0 will come as a shock to many.

I agree. The question is: what will AD 0.5 look like?

 

[182RG]

I think it is highly unlikely that the Model 3 final assembly line in 3Q17 will be mostly robots. Remember Elon referred to this as AD 0.5, and said that AD 1.0 would come a year later. While I do expect more automation than on S/X final assembly line, there will still be many humans working on the Model 3 line in 2017.

^ This ^

3Q17 is 9 months away.

I predict (M3 only)

4Q17 - 250 hand-built cars to "meet the Tweet" and be able to plant the flag.
All of 2018 - 50,000 cars, in an slowly increasing level of assembly automation.
All of 2019 - 150,000 cars.
All of 2020 - 400,000 cars.

I think this ramps a lot slower, with a lot less automation. I don't expect my M3 until 2020. I don't expect one for $35,000 ever.

 

[JRP3]

I think this ramps a lot slower, with a lot less automation. I don't expect an M3 until 2020. I don't expect one for $35,000 ever.

I like the optimism Do you mean you won't see your personal M3 until 2020 and it will cost more than $35K, or that no M3 will ever be sold for $35K?

 

[182RG]

I like the optimism Do you mean you won't see your personal M3 until 2020 and it will cost more than $35K, or that no M3 will ever be sold for $35K?

Thanks!

Both. I reserved early May 2016, 30 days after reservations opened specifically because I didn't want one of the first several hundred thousand. I plan on optioning beyond the base model, to mid $60s.

I don't believe that there will ever be, to anyone, a stripped model that invoices for $35,000.

 

[3Victoria]

1. There will be base models, if only to prove naysayers wrong. I believe the ramp will be much faster with 5-100k by end of 2017.

 

[dennis]

I predict (M3 only)

4Q17 - 250 hand-built cars to "meet the Tweet" and be able to plant the flag.
All of 2018 - 50,000 cars, in an slowly increasing level of assembly automation.
All of 2019 - 150,000 cars.
All of 2020 - 400,000 cars.

I think this ramps a lot slower, with a lot less automation. I don't expect my M3 until 2020. I don't expect one for $35,000 ever.

Even without a lot of additional automation in AD 0.5, I believe Tesla should be able to produce at least double the current S/X quarterly volume with the first Model 3 assembly line. This would mean at least 50K cars/quarter once they get ramped up but before AD 1.0 in mid-2018. The reasons for this:

- current final assembly line builds both S and X so its throughput is limited by the more complicated Model X
- Model S will be simpler to build than both S and X
- Big order backlog allows Tesla to batch builds to optimize manufacturing efficiency
- This will be Tesla's 3rd generation final assembly line in Fremont - they have learned a lot

 

[Sudre]

I don't believe that there will ever be, to anyone, a stripped model that invoices for $35,000.

That is hilarious. The majority of reservation holders I talked to while assisting with traffic and parking at the St. Louis Model ≡ first day event were only going to be able to afford the base model and they would be happy with it. Much like I was one who purchased an early 60kWh S with only air-suspension, leather, as an option and received it in Feb 2013. Others even received their 40kWh S version base model within 2013. This lie that Tesla won't produce and base model cars is FUD for the bears. IF they don't produce any it will only be because no one ordered one. That is highly unlikely since many can barely afford the 35K.

(edit) I am not going to change my opinion just because the car ends up actually costing $35,600 (or whatever) when it comes out... and YES there will be a delivery charge. My point is the cheapest base model around 35K will be delivered if one is ordered.