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

Articles/megaposts by DaveT

This site may earn commission on affiliate links.
Status
Not open for further replies.
I could be wrong, but what I thought I heard was that the cells would hold about 30% more energy, because they would be bigger. But because they are going to be bigger, the energy density of the cells won't (necessarily) be much better. Now, if they're slightly bigger, they might pack better, so the energy density of the battery might actually improve. And yes, new chemistry might also come into an improvement.

Actually assuming the form factor is similar - ie round battery cell, bigger wouldn't pack better - if you mean better packing density.

My sense is that within a given fixed dimension container everything packs better the smaller they go. EG if the diameter of the cells approaches zero - the packing density goes to 100%. Not the other way around. But then the material costs (cathode, anode and external cladding goes up with decrease in battery size). And I don't know the impact on overall energy density.

The not inconsequential energy density increase (30%) quoted for a given fixed carrier (pack level) has to be a result of major battery chemistry improvements and then to a lesser extent battery dimensional tweaks. I just cannot see how any further battery size optimization on current chemistry will yield such a significant percentage.
 
Last edited:
My $0.02 - don't forget that the 18650 is a form factor developed years ago with no particular use in mind - it could be laptop batteries, e-cigarettes, tesla car battery packs, radio-controlled planes or anything else. Panasonic literally has no idea what the end users might do with it.


Tesla packs thousands of these vertically into their battery packs.


Looking at the Tesla pack from above, where each cell appears as a disc... consider the thickness of the cell casing, and the percentage cross-sectional area of that casing compared to the cross-sectional area of the inner battery chemicals. The casing might be 3%, with the latter being 97%. (just wild guesses)


The casing is there to insulate the battery chemicals from whatever environment the battery is in - and to provide it with some strength. Don't forget... the design caters for all environments and especially where the battery is alone - not surrounded by dozens of other identical batteries. Each battery chemical pile is separated by TWO thicknesses of the battery casing.


If Tesla can re-engineer the batteries with a thinner outer casing, it could increase pack density because there is less casing, in total. As far as vehicle performance is concerned (speed, range etc.) - the cases of all those thousands of batteries are nothing but "dead weight." If you can reduce that to the absolute minimum, by logic the rest of the pack must have more functional battery (or copper crossbars, etc. etc.)


Perhaps Tesla has come to determine that a battery with a larger diameter than the 18650 is better, because for a given amount of kWh, there is less casing used - therefore, the battery is denser. The larger air gaps between the cells must act against that curve, so presumably there is a Goldilocks sweet spot where two curves intersect - you have as little casing as possible, but also, the air gaps between cells aren't too large.


The Tesla pack would be much denser if the batteries were hexagonal... not sure that will ever happen (though with the Gigafactory... who knows).


At least air gaps don't weigh anything. Cases have mass, so if you can reduce that, the Wh/Kg is better.
 
#9 Tesla’s Software Advantage (Competitive Advantage #3)

(This is post #9 in a series explaining my long-term TSLA investment philosophy. For previous posts, see Articles/megaposts by DaveT)

Tesla has software chops. In this post I’ll make a case why Tesla’s software expertise is a competitive advantage.

Since Tesla doesn’t need as much mechanical engineering as other ICE auto makers (EVs don’t need an engine or complicated transmission), Tesla is able to use that attention and direct it toward software and electrical engineering.

Let’s look at several areas that Tesla has focused their software development.

1. Battery management
In order to manage the thousands of 18650 battery cells, Tesla has developed an advanced system of thermal management that relies on software. I look at the Tesla battery pack as a computer that monitors each individual cell and makes sure they don’t overheat, are used efficiency, and are charged safely.

Prior to Tesla releasing the Model S many people thought it was impossible for Tesla to use 18650 cells in a manner that would be safe and efficient. For Tesla they were able to choose the cheapest and highest energy density form factor available and overcome the challenges of safety and cell management through their software implementation.

Since other companies don’t have the software expertise of Tesla, they tend to look at the 18650 implementation for an auto battery pack as overcomplicated and not worth the risks. But for Tesla, software is one of the main reasons why they were able to release the Model S with the range it has and at a competitive price point.

2. New features via software updates
Now some people think of Tesla’s software updates as just infotainment updates. However, if one looks deeper they’ll discover that Tesla can update quite a few features via software updates and it shows how integrated and advanced the car is.

One example is Tesla’s release of hill hold via a software update several months ago. Adding hill hold is not just a simple infotainment feature. Hill hold (where the car remains in braked position for a second before pressing the accelerator when on an incline) is very advanced feature to add since it deals with the core accelerating and braking features of the car. In other words, via a software update Tesla was somehow able to add a feature where it can detect if the car is on an incline (perhaps this was already being monitored) and then if the car is on an incline to keep applying brake pressure (or somehow keep it from going backward down the hill) for a second before the accelerator is pressed.

The hill hold feature shows me that Tesla can update core driving functions (ie., accelerator, brake, etc) via software updates. This is interesting since from a safety standpoint, typically one would design the car where the core driving functions can’t be touched/updated by an over-the-air software update since it introduces a lot of complexity and potential safety concerns. But somehow Tesla’s intent was for the Model S to be updated not just with infotainment features but also for software updates to adjust even core driving features. This is quite remarkable, and probably something that other auto markers will have a difficult time justifying since they don’t have the software expertise and it’s probably not worth the risks to them.

Another example of a new software feature was the ability to adjust air suspension at different speeds. For example, a software update a few months ago allowed users to keep the Model S at the lowest air suspension setting at all speeds or the owner could decide at what speed they want the suspension to lower. This is another example of how software updates aren’t just updating infotainment features but are updating core driving features, including air suspension settings.

This is not to discount Tesla’s implementation of their infotainment center console, which I think is roughly 5 years ahead of the competition. However, adding and modifying core driving features via software updates is in a whole other league and existing auto makers will likely find this one very challenging to match. It just goes to illustrate Tesla’s software advantage.

3. Continuous Feedback
Another brilliant move by Tesla was the release of the connected car (ie., the car always having an internet connection). This allows the car to give feedback to Tesla (ie., # miles, major problems, etc). It gives Tesla a sort of continuous feedback that Tesla can use to figure out where reliability issues are (in conjunction with info from service centers, etc) and try to fix them quickly.

When a car pulls up and connects to a Supercharger, I’m also guessing that Tesla is gathering information on how the cars are charging and their state of charge, and using that information to improve Supercharging. Again, it’s continuous feedback. Usually with a typical car, you drive it off the dealer lot and it’s sayonara to the dealer and car company until you take it in for service some day in the future. However, for Tesla it’s different. They actually want continuous feedback, and they’ve built it into the system so that can use the data to make improvements. Another example of this is when a Model S is involved in a serious accident it appears that the car alerts Tesla and Tesla is able to reach out to the owner to see if they can be of assistance but also so Tesla can investigate the cause of the accident, if needed. This allows Tesla to quickly fix any potential massively dangerous issues.

Many tech and software companies today understand the important of a tight and quick feedback loop. By having a tight feedback loop, companies can quickly identity/solve problems and they can iterate to make their products better. Often a tight feedback loop is the difference between a company succeeding and failing.

Another area of continuous feedback is their stores. Tesla is able to monitor how many people visit a store each day, what kind of questions are asked, how many test drives are taken, how many orders are placed, etc. The reason this is important is because Tesla can have a real-time pulse on demand and their potential customers.

A growing field in the tech/startup world is growth hacking, which is simply trying to tweak processes and design to improve growth rates. In order for growth hacking to work, one makes minor change and then views the results to see if those changes made any difference. The key is having a quick feedback loop that makes the results obvious. In Tesla’s case, they’re able to do growth hacking (ie., adjust processes, etc) and see the immediate results via their stores. For example, in the future if Tesla wanted to try an advertising campaign they could try a few versions. Version #1 would go out this week and they could track if store visits increase or not. Version #2 would go out next week and they could see if store visits increased or not (compared to Version #1). And the following week Version #3 would go out and store visits could be compared with previous weeks. Also, different approaches possible where you release Version #1 in region A in week one and Version #2 in Region B in week one, and compare the results via store traffic and orders placed in those regions. It’s another advantage of continuous feedback.

Another area of Tesla’s software expertise is their internal software. Tesla is able to accept orders online and then follow the customer’s entire experience through purchase and post-purchase. Also, Tesla has developed their own software to manage many internal processes as well. This will help them become more efficient over time.

Now I’m not saying that other auto makers won’t be able to add some software expertise to their companies. They are already doing this. However, the question is whether existing auto makers will be able to move as fast as a software company in the field of software. In order to do so, existing auto makers are going to have to radically restructure their companies and I don’t see that happening.

Tesla’s advantage is that their CEO understands software (his first two startups were software startups) and they’re headquarted in an area where many of the best software engineers live. As a result, Elon and his team has infused Tesla with a value and ethos to respect and integrate software deeply into their company. This value and ethos simply doesn’t exist at existing auto companies. And as a result, existing auto companies will likely be playing catch up to Tesla in the area of software for many years to come.
 
Last edited:
Hi Dave,

I just watched the most recent Google Hangout discussion. Great stuff. I wanted to point out something that I felt was overlooked.

The group speculated on the battery improvements at length. For some reason, I haven't heard any investor mention that 2 years have passed since the introduction of the Model S and therefore 2 years since the first Model S battery pack.

​Given the approximate 7% annual rate of battery improvement, the batteries in the today's Model S are most certainly superior to the batteries in last year's (or the year before) Model S. And if the batteries aren't superior, then they are most certainly cheaper for Panasonic to make, and therefore cheaper for Tesla to buy.

I also feel comfortable saying that the ever-so-slightly improved batteries have contributed to the reduced weight of the Model S.

Too bad no one cared to verify this on the earnings call :)

As for the $200-300 per kWh you mentioned, I expect that we are already in the lower end of that range.
 
Good to recognize their software skills however I think a big one being missed is their internal production and planning systems. A while ago there was mention of how they built their own enterprise resource planning system which, by itself, is pretty impressive ...but really impressive if you marry that concept with their continuous product improvement cycle. Personally I think they have a system that tracks the version of every part that goes into every individual car so they can change / upgrade a part at any time and have it tracked all the way until you go to service.

This is has a lot of production implications of course, but also some significant ones regarding service and eventually third party integration

huge production advantage
 
Good to recognize their software skills however I think a big one being missed is their internal production and planning systems. A while ago there was mention of how they built their own enterprise resource planning system which, by itself, is pretty impressive ...but really impressive if you marry that concept with their continuous product improvement cycle. Personally I think they have a system that tracks the version of every part that goes into every individual car so they can change / upgrade a part at any time and have it tracked all the way until you go to service.

This is has a lot of production implications of course, but also some significant ones regarding service and eventually third party integration

huge production advantage

Very perceptive 30
That comes from the SpaceX integration which requires 100% traceability on all components to meet NASA requirements. His includes trace to raw components, and feeds to the GF deployment that couples vertical integration to raw material.
 
#10 Product Excellence (Tesla’s Competitive Advantage #4 part 1)

(This is post #10 in a series explaining my long-term TSLA investment philosophy. For previous posts, see Articles/megaposts by DaveT)

In this post I’ll dive into product excellence as a competitive advantage of Tesla.

Tesla’s product excellence begins with high standards. This is probably the attribute that distinguishes Tesla the most and is best seen in Elon’s unrelenting focus to deliver a stellar product in all aspects.

It’s difficult to exaggerate how good of a car the Model S is. It handles well, accelerates fast, looks good, has ton of cargo space, great range, can go across country with supercharging, and can even get better with over-the-air software upgrades or eventual battery upgrades/replacements.

However, in order to make the Model S as good as it is, Tesla had to go through a process where they looked at almost every aspect of the car and had to ask themselves how it could be better and come up with creative solutions to bring each aspect of the car closer to ideal.

Design Thinking

In recent years IDEO and other companies have pioneered and propagated a concept called Design Thinking. Design Thinking is where one uses creativity and out-of-the-box thinking to solve problems and make products better. A good example is the company OXO and their kitchen products. Oxo has reinvented dozens if not hundreds of kitchen utensil and food prep items. They take a typical potato peeler, as an example, and ask if a typical potato peeler is peeling a potato in the best possible way or if there’s anything they could do to redesign the potato peeler to make peeling potatoes better. This is a surprisingly arduous process that takes a ton of brainstorming, iteration and experimenting. Eventually, with a lot of focus and execution, the result is the potato peeler redefined. Or simply a better potato peeler.

Another example of design thinking is the SimpleHuman kitchen trash can I bought a few years ago (Amazon.com - simplehuman Rectangular Sensor Trash Can, Fingerprint-Proof Brushed Stainless Steel, 40-Liter /10.5-Gallon - Simple Human Bin). It costed $200 at that time, and my wife was very skeptical of the purchase. She was like, “A $200 trash can!?!” But the trash can is quite marvelous. They really re-designed the trash can and made it a lot better than any other trash can I’ve used. The sensor just works, the trash is easy to pull out, the lid stays up when you replace the trash liner, the trash liner fits perfectly, etc. Everything just works and works better.

Bringing this back to Tesla, Tesla has enacted Design Thinking (creativity, user empathy, out-of-box thinking, problem solving, etc) to redesign as many concepts/functions while making the Model S. They took out all the buttons and replaced it with a huge touchscreen that can be updated over-the-air. They made the car fast with long range. They made a car that looks great but has a super low drag coefficient. They simplified the driving experience by allowing the driver to start the car by simply pressing the break and pulling a lever. They reinvented the car purchase and order experience. They reinvented the long-distance charging experience for electric vehicles. They are trying to reinvent the service experience. And there’s a ton of other aspects where they’re trying to reinvent but we don’t have access to (ie., internal processes, production methods, insourcing products, etc).

User Empathy

Elon and Tesla remind me of the user empathy found in some of the greatest companies/innovators of our time. Steve Jobs had a deep user empathy where he detested things that would complicate and overwhelm the typical user, thus the iPhone/iPad is remarkable simple to use. Google had a deep user empathy where they wanted to deliver the best and fastest search results to the user, while other search companies at the time (in the early days) cluttered their home pages with a million links and didn’t care much about the quality of the search results.

Often, companies lose sight as they focus more and more on making money, bringing in revenue, achieving growth, and delivering on earnings. The customer becomes just a means to reach a profit/revenue goal.

However, for visionaries like Steve Jobs and Elon Musk, the company only exists to bring value and make the lives of people better, and to the extent that they can accomplish that the companies will be rewarded financially, although financial goals aren’t the most important goal. Rather, the more important goal is to make the lives of people and the world better by providing amazing products/services that are far better and deliver greater value than what’s already out there.

This value of user empathy is not easily copied or imitated, since it’s a deep internal value. What Tesla has going for it is it’s been inseminated with this value of deep user empathy from Elon and his team, and this value guides the company to not only make the best products/services they can but also to make the right decisions going forward.

Improving Quality

Another aspect of product excellence is the way Tesla is continuously improving the Model S. Tesla has a strong control over what the reported issues customers face, and they have their team of engineers involved in addressing and finding solutions.

In the most recent quarterly conference call Elon shared details on his weekly Product Excellence meeting:
“Every week I have a product excellence meeting, which is a cross-functional group, so we've got engineering, service and production, and we go about all the issues that customers are reporting with the car, and develop action items to address to get car ultimately to the platonic ideal of a perfect car - that's what we're aiming for. Because although I think we've got great service, the best service is no service. That's really what we want, is a car that never needs to be serviced. And I think we're getting there quite rapidly.”

He also shared his aspiration to have the Model S (and other Tesla cars) have a quality unmatched by other cars:
“I mean our aspiration is sort of order magnitude better quality than any other car. And we'll keep at it unrelentingly until we get there.

It’s impressive that Tesla’s CEO is hands-on in addressing quality control issues and he exudes a vision of quality that is quite inspiring. On top of this, Tesla is able to assemble a cross-functional group to lead this effort since Tesla owns all their service centers and is also an engineering-driven company. This allows engineers the freedom to innovate and iterate on solutions, then for service centers to try those solutions out, and if the solution work then for production to immediately incorporate the changes into the production line.

The key to all of this is the speed at which a company can address a quality control issue. By bringing in key cross-functional personnel under the direction of the CEO, Tesla is able to quickly address issues that crop up. The result is the car gets better and better at a faster rate than any other car on the market.

Owners love the car

For a year and a half I followed the used car market closely and I would regularly email and talk to Model S owners who were selling their car. I wanted to know why they were selling their car and what their next car would be. While I did meet some owners who preferred other cars over the Model S and were selling their Model S as a result, the vast majority of the used Model S sellers were people who loved their cars and were getting a bigger battery, a performance model, or buying a new car with different options/colors/etc.

The general consensus among Model S owners is that they are very, very happy with their car. All of the effort that Tesla has invested in Design Thinking is paying off as owners appreciate the range, performance, supercharging, cabin space, technology, and the electric drive.

The result is strong and growing demand that allows Tesla as a company to be even more ambitious in their future growth plans.

A few final notes:

1. If you haven’t seen this video, this is a great one to watch. It’s the Consumer Reports auto team reviewing the Model S and sharing their thoughts (back in May 2013). Note how enthusiastic and impressed they are at how Tesla has rethought so many aspects of the car. Talking Cars with Consumer Reports #5: Tesla Model S - YouTube

2. The Model X. Elon and his team is going over the Model X with a fine-toothed comb (TMC Connect Snippets / Gems). The Model X exemplifies everything I’ve written here about high standards, design thinking and user empathy. It’s going to be a great car that redefines what an SUV can and ought to be.

3. Gen3. Elon understands how important product excellence is to the success of Tesla’s Gen3 vehicle. This is the main reason why he’s committed to Tesla as the CEO until Gen3 enters high-volume production. It’s because he knows that he needs to make sure the Gen3 is near-perfect and exudes the high standards and design thinking that he knows is needed to make it a success.
 
Hi Dave,

I just watched the most recent Google Hangout discussion. Great stuff. I wanted to point out something that I felt was overlooked.

The group speculated on the battery improvements at length. For some reason, I haven't heard any investor mention that 2 years have passed since the introduction of the Model S and therefore 2 years since the first Model S battery pack.

​Given the approximate 7% annual rate of battery improvement, the batteries in the today's Model S are most certainly superior to the batteries in last year's (or the year before) Model S. And if the batteries aren't superior, then they are most certainly cheaper for Panasonic to make, and therefore cheaper for Tesla to buy.

I was trying to cover this a bit during the call.

The 7% annual rate is something that has been true over time. But each improvement still requires specific breakthroughs to make them happen. It's only when you zoom way out that the 7% annual rate is true.

Therefore, I prefer to look specifically at what the breakthroughs might be to make that 7% true and that gives us a better sense of the short to medium term timing. Assuming that the Model S is already using a cell that is mostly based on the Panasonic NCR18650A or NCR18650B, the next logical step is the silicon anode 4.0Ah cell. However, we have not seen that on the market yet, even though Panasonic issues some press releases and has subsequently blown the release dates that were given in those press releases. The roadmap did pick up a 3.6Ah cell in the meantime, but it isn't clear how they get to 3.6Ah. Note that the silicon anode is heavier, so the improvement in specific energy is a smaller gain. I had hoped that the Model S was based on the NCR18650A since it would mean that they might be able to apply the technology in the NCR18650B to the Tesla cell and get a 8-9% specific energy gain. JRP3 has mostly disabused me of that notion and it is quite likely that the current Tesla cell is already using NCR18650B technology.

As for the price, there was a Taiwan industry group report on lithium ion pricing that put 18650 lithium ion cells at a $150-$200 per kWh price point in Q4 2012. That's wholesale pricing for the CE industry. One can assume that Panasonic's top of the line NCA cells are at the top of that range, but Tesla is also probably paying a substantial volume discount. Mr. Musk has also references hitting production milestones in 2013 that lowered cell costs. Further, they negotiated a new cell purchase agreement last fall. Given the retail pricing of around $270/kWh at the battery pack level and a GM around 25%, the cell costs are probably under $200/kWh right now, but battery costs somewhere just over $200/kWh. To be conservative, I would model $220/kWh at the battery pack level.

Now, you might say, well, you can't buy a Model S battery from Tesla at $18,700 x 1.25 = ~$24,000. Right... They are marking it up an additional almost 100% because they are cell constrained. Really the only entities really buying Model S battery packs are insurance companies and each cell sold for repair is a cell that can't be sold as part of a car and therefore profit lost. I think they are charging the entire profit of a car onto each replacement battery pack right now which goes for about $46,000 (85kWh), or a difference of $22,000 which roughly translates to the 25% GM markup on the ~$100,000 ASP.

Most interesting during the call is that Tesla expects to ship a new cell chemistry and a new form factor. There are all sorts of implications for that. For a while there, I was surmising that it is possible that Tesla does not change the cell at all and goes for the absolute lowest price per kWh. The current cell is "good enough" and if the price was, say, $135/kWh, there would be significant disruption without the need for any chemistry or form factor changes. But since they are building the Gigafactory, they have the luxury of doing all of it - new form factor, new cell chemistry and lower pricing.

I think if we look at the specs of the announced but not shipped 4.0Ah Panasonic cell, we can get a sense for what Tesla is going to ship, but we have to account for the changes in form factor.

This also begs the question, will the new cells fit in the current Model S pack design or will we need a change to the Model S design at that point? Also, will Panasonic be shipping the new cell or the old cell design out of Osaka, even after the Gigafactory is online?
 
The 7% annual rate is something that has been true over time. But each improvement still requires specific breakthroughs to make them happen. It's only when you zoom way out that the 7% annual rate is true.
There are many paths to this trend: many small improvements, or the occasional step-change. A 23% improvement every third year, e.g., gives a 7% annual rate.

In the case of batteries, I think we're seeing a combination of incremental improvements punctuated by the occasional step-change.
 
DaveT,
this is a great series you keep posting, keep them coming.
a while black some speculated that tesla might do something with the military...do you think that is still in the cards?

I believe doing something with the military is one of the most likely 'we are not showing all our cards' possibilities.
 
DaveT,
this is a great series you keep posting, keep them coming.

It's funny a couple days ago I spent my entire Sunday afternoon writing the next two posts in the series and then my TextEdit program crashed. Usually I just open it back up and everything's restored to where it was. But this time, it was all gone. And I couldn't find a way to recover it. So, I gave up and told my wife in exasperation that I should just stop writing these posts (since who reads these posts anyways, right?). Then, my wife was like "No, we've got to find a way to recover that file." I told her, "Forget it. It's gone. Maybe this is a sign I should take a break from writing on TMC." But she wouldn't give up and eventually we found a way to restore the file via Time Machine. And that's the story how my wife saved this series.
 
It's funny a couple days ago I spent my entire Sunday afternoon writing the next two posts in the series and then my TextEdit program crashed. Usually I just open it back up and everything's restored to where it was. But this time, it was all gone. And I couldn't find a way to recover it. So, I gave up and told my wife in exasperation that I should just stop writing these posts (since who reads these posts anyways, right?). Then, my wife was like "No, we've got to find a way to recover that file." I told her, "Forget it. It's gone. Maybe this is a sign I should take a break from writing on TMC." But she wouldn't give up and eventually we found a way to restore the file via Time Machine. And that's the story how my wife saved this series.

Thanks Mrs. Dave T.

Dave T: Your series is very informative and I believe read by many TMC members (and possibly some analysts who are letting you do their job for them). Usually your logic and attention to detail are so good that no one has any comments....appearing to you perhaps that no one is reading them. Personally, I would be disappointed if you stopped your analysis. Please keep up the good work.
 
It's funny a couple days ago I spent my entire Sunday afternoon writing the next two posts in the series and then my TextEdit program crashed. Usually I just open it back up and everything's restored to where it was. But this time, it was all gone. And I couldn't find a way to recover it. So, I gave up and told my wife in exasperation that I should just stop writing these posts (since who reads these posts anyways, right?). Then, my wife was like "No, we've got to find a way to recover that file." I told her, "Forget it. It's gone. Maybe this is a sign I should take a break from writing on TMC." But she wouldn't give up and eventually we found a way to restore the file via Time Machine. And that's the story how my wife saved this series.

Your posts made me to consider investing in TSLA from India and I don't even trade Indian stocks. May be I did not find such detailed articles on any Indian company.
 
Status
Not open for further replies.