2017 Investor Roundtable:General Discussion

[MitchJi]

Great; thank you.

So now we're at 1MWh for 500 miles for a Tesla Semi (assuming drag coefficients are the same), or ~2kWh per mile.

This also means Tesla Semi's battery size is unlikely to be above 1 MWh.

My WAG is that the semi's will be designed to accommodate up to 4-5 powerpacks that can easily be swapped a long trips. The supercharger stations will be using powerpacks. The main challenge will be the logistics of maintaining the correct number of charged powerpacks at the necessary locations.

 

[ValueAnalyst]

My WAG is that the semi's will be designed to accommodate up to 4-5 powerpacks that can easily be swapped a long trips. The supercharger stations will be using powerpacks. The main challenge will be the logistics of maintaining the correct number of charged powerpacks at the necessary locations.

Why would carrying around powerpacks be a better solution vs batteries? Aren't they essentially the same thing anyway (I.e. Energy storage)?

 

[ChrisA]

Quick question: I think Tesla management has always talked about battery costs in terms of pack costs, and even disparaged others who cite cell costs alone as "battery costs." Thus when Straubel said at the Edison Electric Institute talk (

see minute 36) that he would be "disappointed if their battery costs were not below $100 KWh by 2020", I assume he was referring to pack costs, not cells cost alone. However, he was not explicit about pack costs vs. cells costs in that particular comment. Does any one else have a link that clearly indicates they prefer to talk about pack costs and not cell costs? Or, even better a comment that explicitly shows they are aiming to get below $100 KWh for their pack costs by 2020.

Or, am I mistaken. Was Straubel actually referring to cell costs there?

Any thoughts appreciated.

 

[erthquake]

My WAG is that the semi's will be designed to accommodate up to 4-5 powerpacks that can easily be swapped a long trips. The supercharger stations will be using powerpacks. The main challenge will be the logistics of maintaining the correct number of charged powerpacks at the necessary locations.

Interesting idea, but the form factor might be a problem. It would raise the center of gravity too high unless you could somehow lay them on their sides and slide them under the cab.

 

[MitchJi]

“This is not the first nor the last quarterly Tesla disappointment or miss,” Janus Henderson portfolio manager David Chung told Reuters. “I think longer-term investors like us recognize that we are still early in this sector and this is a truly special company and exceptional CEO.”

https://www.bizjournals.com/sanjose...-stock-fund-manager-short-sellers-profit.html

"In terms of conviction, they seem to be very strong because they are taking it on the chin and keep on going on," said Ihor Dusaniwsky, head of research at S3 Partners, a financial analytics firm in New York.

Graham Tanaka, portfolio manager of the New York-based Tanaka Growth fund, said Tesla short sellers were not factoring in potential improvements and growth of the battery division. The company began production of Model 3 batteries at its Gigafactory in Nevada in January, which should reduce costs and improve the lifespan of batteries with greater scale of production, Tanaka said.

"They're going to be making better batteries at a cheaper price than any of the competition," he said.

 

[MitchJi]

Why would carrying around powerpacks be a better solution vs batteries? Aren't they essentially the same thing anyway (I.e. Energy storage)?

Modular, only carry what they need.

The customers would not need to pay attention the batteries as part of the purchase price.

Easily swapable.

They would be useful at the supercharger stations (assuming that they can make the logistics work).

Just because this makes sense doesn't mean that I'm correct. Tesla could have another plan that's better.

 

[tander]

Has anybody been paying close attention to auto pilot progression? If I recall correctly they said maybe a year or year and a half ago ( can't remember the quote) that they needed hundreds of millions or over 1,000,000,000 miles driven to be able to demonstrate auto pilot as being significantly safer than people even in the weird corner problems. My understanding is once they had enough miles they would download the full auto pilot to the S and x first. Shouldn't that be happening about now?

 

[SwTslaGrl]

From today's Electrek, the linked article in Automobil Produktion - (the VW 2019-2020 perspective?)

"In this optimistic outlook, the expectation of significantly decreasing battery prices is likely to be taken into account. As Sedran said, there was considerable progress in terms of both price and performance. For lithium-ion batteries Sedran expects a drop in prices from currently 150 - 200 euros per kilowatt hour to under 100 euros."
Sedran: VW I.D. 7.000 – 8.000 US-Dollar billiger als Model 3

 

[FutureTSLACEO]

Has anybody been paying close attention to auto pilot progression? If I recall correctly they said maybe a year or year and a half ago ( can't remember the quote) that they needed hundreds of millions or over 1,000,000,000 miles driven to be able to demonstrate auto pilot as being significantly safer than people even in the weird corner problems. My understanding is once they had enough miles they would download the full auto pilot to the S and x first. Shouldn't that be happening about now?

Elon Musk Says Autopilot Needs 6 Billion Miles Before Regulations | Inverse

Now at 5 billion miles.
Tesla’s global fleet reaches over 5 billion electric miles driven ahead of Model 3 launch

Hopefully, 6 billion miles by the end of the year. Full autopilot driving may be done by the end of the year IMO. In fact, Elon said a demonstration will be done using a Model S from NY to CA by the end of this year. Elon Musk says a Model S will drive itself from New York to LA this year

TSLA touching 450 IMO by then if macros and guidance are in line. (Not an advice)

 

[MitchJi]

Interesting idea, but the form factor might be a problem. It would raise the center of gravity too high unless you could somehow lay them on their sides and slide them under the cab.

I don't believe that's a potential problem.

 

[MitchJi]

Hopefully, 6 billion miles by the end of the year. Full autopilot driving may be done by the end of the year IMO. In fact, Elon said a demonstration will be done using a Model S from NY to CA by the end of this year. Elon Musk says a Model S will drive itself from New York to LA this year

That doesn't mean that FSD is "done".

 

[ValueAnalyst]

Elon Musk Says Autopilot Needs 6 Billion Miles Before Regulations | Inverse

Now at 5 billion miles.
Tesla’s global fleet reaches over 5 billion electric miles driven ahead of Model 3 launch

Hopefully, 6 billion miles by the end of the year. Full autopilot driving may be done by the end of the year IMO. In fact, Elon said a demonstration will be done using a Model S from NY to CA by the end of this year. Elon Musk says a Model S will drive itself from New York to LA this year

TSLA touching 450 IMO by then if macros and guidance are in line. (Not an advice)

I think he meant 6 billon AUTOPILOT miles, but the 5 billion mile figure you cited I believe is total miles driven by all Tesla's (autopilot or not).

So I would say give it another year for the 6 billion Autopilot miles to be accumulated.

 

[adiggs]

Elon Musk Says Autopilot Needs 6 Billion Miles Before Regulations | Inverse

Now at 5 billion miles.
Tesla’s global fleet reaches over 5 billion electric miles driven ahead of Model 3 launch

Hopefully, 6 billion miles by the end of the year. Full autopilot driving may be done by the end of the year IMO. In fact, Elon said a demonstration will be done using a Model S from NY to CA by the end of this year. Elon Musk says a Model S will drive itself from New York to LA this year

TSLA touching 450 IMO by then if macros and guidance are in line. (Not an advice)

I like the optimism, and having used autopilot for a 60 mile drive during an overnight test drive, I'm all in favor of sooner than later for autopilot to be a reality. (It made for a big reduction in the fatigue from the drive).

However the two cited articles have two different (though related) ideas. The first is a year go, Elon thinking that it would take 6 billion miles on autopilot to achieve worldwide regulatory approval.

The second article is total miles driven by all Tesla vehicles. Unclear from the article, but I think the 42k miles on my Roadster odometer would be included in those miles (along with the mileage for all Roadsters). Pretty clearly to me from the second article, all of the pre-autopilot Model S' are included in those miles.


Putting the two articles next to each other implies that we are 83% of the way to full self driving with worldwide regulatory approval. I claim that is misleading. Even if we fudge the Tesla fleet miles down to 3B of autopilot miles, that would imply that we're 50% of the way to worldwide regulatory approval (I don't believe that, but I could be wrong of course).


My own guess is that we've got at least another year for the first regulatory district to provide approval for full self driving (vehicle in control and responsible, no driver in car).

I consider that to really be too aggressive, with another 1-3 years of lots of companies doing lots of testing and pilot projects, with cars becoming increasingly autonomous but with humans sitting behind the wheel and monitoring. Even the cross country demonstration will be exactly that - a technology demonstration. It'll be really cool for them to get into the car in LA and have it park itself at each supercharger along the way (wait for human to plug it in and then unplug it). The car will still have a human behind the steering wheel, even if they don't do anything for several days while driving cross country.

There's still a LOT of work to go from LA to NY technology demonstration completed one time, to worldwide regulatory approval.


My guess - 60 Billion autopilot miles isn't enough for worldwide regulatory approval. 600 Billion might not be enough, but worldwide regulatory approval is a pretty high bar to clear.

The good news for all of us as Tesla investors is that nobody is accumulating miles at anything like the pace Tesla is. Considering the mechanism Tesla has in place, we WANT the problem to take 600 Billion miles or even 6 Trillion miles. Because the only company with a solution in place to economically accumulate that level of training data / experience is Tesla. If that's the standard, then nobody else has yet started on an economic path to accumulate that level of machine experience.

 

[RobStark]

What do you think the effect of a $3-$6k price reduction would have on demand, after the M3 is launched?

Offer a Model X 75 with rear wheel drive, standard coil suspension, and five seats for $74,500. Let customers pick and chose what options they want. IMO this would boost demand for X by ~20-25%.

There are plenty of soccer moms outside the snowbelt that prefer CUVs for their interior size and seating position without the stigma of being Minivans/MPVs. They don't need and prefer not to pay for AWD active suspensions.

 

[elasalle]

Southern Company building the future of energy with new battery storage research demonstration

ATLANTA, July 12, 2017 /PRNewswire/

Located in Pensacola, Florida, at Gulf Power's Douglas L. McCrary Training and Storm Center, the new research project will test and evaluate a 250 kilowatt/1 megawatt-hour Tesla Powerpack lithium-ion industrial energy storage system over a two-year period. Insights gained from the demonstration are expected to accelerate the development of battery storage technology across the Southern Company system


Glad to know interesting things happening now in Florida as well. Till like last Nov, SolarCity did not even have a presence there - due to all the lobbying from Utilities etc.

 

[Reciprocity]

Offer a Model X 75 with rear wheel drive, standard coil suspension, and five seats for $74,500. Let customers pick and chose what options they want. IMO this would boost demand for X by ~20-25%.

There are plenty of soccer moms outside the snowbelt that prefer CUVs for their interior size and seating position without the stigma of being Minivans/MPVs. They don't need and prefer not to pay for AWD active suspensions.

Agreed, but I dont think they will. In part, I dont think they can build them fast enough and I think they want the highest possible margin if they are supply constrained. I think if they do what you are suggesting, its a sure sign that things are not good on the demand side or they have found a way to make a lot more of them.

 

[MitchJi]

Has anybody been paying close attention to auto pilot progression? If I recall correctly they said maybe a year or year and a half ago ( can't remember the quote) that they needed hundreds of millions or over 1,000,000,000 miles driven to be able to demonstrate auto pilot as being significantly safer than people even in the weird corner problems.

My understanding is once they had enough miles they would download the full auto pilot to the S and x first. Shouldn't that be happening about now?

Three separate issues.

1. They need 6 billion miles to establish the fact that their FSD is safer than a human driver.

2. Do you believe that they can start doing that before they have at least completed their LA-NYC demo?

3. The current bottleneck is clearly not getting regulatory approval. The current bottleneck is obviously finishing getting the SW to the point where it is "significantly safer than people" in normal driving, as opposed to "weird corner problems". You need to learn to walk before you can start to run.

 

[Intl Professor]

Why are you trying to kill sex?

My skirmishes even predate the song: "Footsteps on the dashboard, upside down." I do remember the melody. But the malady doesn't linger on. (Apologies to Bob Hope; may he continue in memories.)

 

[Drumheller]

I'm only one opinion, and I'll probably be disliked for this post, but I think Level 5 is a long way off. Years. Level 4 is probably some years away. I've been in software dev for over 20 years and I don't think this is an easy problem, at all. There are a huge number of edge cases that, due to safety, must be accounted for before true FSD is ready.

I'd like to be wrong.

 

[ValueAnalyst]

I'm only one opinion, and I'll probably be disliked for this post, but I think Level 5 is a long way off. Years. Level 4 is probably some years away. I've been in software dev for over 20 years and I don't think this is an easy problem, at all. There are a huge number of edge cases that, due to safety, must be accounted for before true FSD is ready.

I'd like to be wrong.

Just wanted to grant you your wish. You're welcome.