Model 3 [Official] Tesla Model 3

[Cashmere]

No, certainly not. When it comes to autonomous functions, for instance, both Daimler and Volvo are still ahead. Even Audi in some aspects.

I appreciate your response but you did not provide a single explanation as to why Audi/Daimler/Volvo are ahead. I will try and explain why I think Tesla is leading the self driving race, and i'll summarize it into 4 categories:

1) Developing autonomy on ICE vs EV

Developing a self driving car is very difficult. Developing a self driving car on an ICE platform is VERY DIFFICULT x 100 - even more lines of code need to be written to communicate the self driving software with the transmission, air/fuel flow, and the thousand other parts that go into an engine. Now, if that is not complicated enough, you can add an extra layer of complexity by introducing hybrids (ICE + battery + electric motor). Its an even BIGGER cluster-fu*k. Making hybrids in the early 2000's made sense to familiarize car makers with how batteries and electric motors work together while they gradually phase out ICE. Making hybrids in this decade is regressive. They add an extra layer of complexity to everything from manufacturing cost/process to repairs - solely for the reason of showing pseudo emission numbers. They don't move the industry forward - they just tick a box in a sandbox test that says "when X car is driven at Y speed it emits less than Z emissions".

Add to that, the current fuel station infrastructure is not designed for automatic refuelling. A human is needed to pump the fuel into the car. So for cross country travel, all petrol stations would have to be upgraded. This is another anomaly beyond the control of ICE carmakers as they would have to beg their brethren at the OPEC Cartel to make upgrades.

These are obstacles that Tesla does not face. The simplicity of EV cars (far fewer parts) means that Tesla's pace of software development is leaner and significantly faster. Tesla is also vertically integrated with its own supercharging network. If they decide to incorporate wireless charging at all their stations tomorrow, they can easily do so without needing to wait for anyone.

2) Democratizing Technology

Chapter 1, Page 1 out of every successful tech startup's guide to success is making your technology scalable. This usually entails suffering losses in the beginning but once you scale upwards, thats when you begin to reap the benefits. Last year September when Tesla showed off Autopilot 2.0 with the 8 cameras they showed the following video:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

If you watch the video, you can observe how camera is capable of discerning between lane lines, in-path objects, out of path objects, road signs, road lights and motion. The amount of data 1 camera captures is staggering. With 8 x cameras + radar & ultrasonic sensors you are crunching through teraflops of data. To crunch through that kind of data you need a supercomputer.

Having these array of sensors paired with a super computer on a $70,000 car is impressive. Having it on a on a $35,000 car is disruptive. Do you think the Germans will ship this kind of hardware on a base C Class, A4, 3 series? Highly unlikely because they're set in their old ways of making high profit margins and reserving cutting edge tech for the S Class/A8/7 Series and only updating their cars once every 5-10 years. Even if they miraculously decide to ship this type of hardware on all their cars from next year, do you think they would be able to sell it for $35K and still make a profit? The BMW 320i costs $34K and they still make you pay extra for a reverse camera.

3) Size of Dataset:

You may have heard, data is the new oil?

If you were to take all the self driving cars from Google, Mercedes Benz, Volvo, BMW, Audi, Volkswagen and add them together, you would find that they probably have less than a thousand test cars running around in their self driving fleets. To be generous, lets just assume they have 5,000 cars.

So far in 2017, Tesla has already sold +50,000 cars with all of the above hardware. Tesla already has a fleet of test cars that is 10 fold bigger than all of their rivals put together! By the end of 2018 (if the are able to deliver Model 3s), there will be +500,000. The company with the biggest dataset will also have the most refined algorithm for self driving cars. With every car Tesla sells, they're putting an exponential gap between them and every other carmaker that doesn't sell a car with self driving hardware.

4) Legislation and shadow mode

When Elon Musk introduced AP2 last year, he spoke about Shadow mode.

Before Teslas can start driving autonomously, the company needs to collect a lot of data to prove to customers (and regulators) that the technology is safe and reliable. So, the car will run Autopilot in “shadow mode” in order for Tesla to gather statistical data to show false positives and false negatives of the software. In shadow mode, the car isn’t taking any action, but it registers when it would have taken action. Then, if the Tesla is in an accident, the company can see if the autonomous mode would have avoided the accident (or the other way around, with the self-driving system potentially causing an accident). It will record how the car would have acted if the computer was in control, including information about how the car might have avoided an accident (or caused one). That data would then be used to show “a material improvement in the accident rate over manually driven cars,”

With Shadow mode and a huge fleet of cars, Tesla is testing and improving self driving cars at a scale that no other car maker is. Once they have substantial data, Tesla will be able to go to legislators and show them a logbook of all the data and how AP is safer than normal driving. Once a computer can carry out a task more efficiently and more safely then a human, then there will be no reason for it to not get approved.

The adoption of self driving cars will happen far quicker then many people expect. While the rest of the car industry is trying to figure out how to make affordable electric cars, Tesla's secret weapon is actually its lead on self driving capabilities. IMO, Model 3 is arguably the most important car to come around since the birth of the automobile.

 

[Centurion]

4) Legislation and shadow mode

When Elon Musk introduced AP2 last year, he spoke about Shadow mode.

Before Teslas can start driving autonomously, the company needs to collect a lot of data to prove to customers (and regulators) that the technology is safe and reliable. So, the car will run Autopilot in “shadow mode” in order for Tesla to gather statistical data to show false positives and false negatives of the software. In shadow mode, the car isn’t taking any action, but it registers when it would have taken action. Then, if the Tesla is in an accident, the company can see if the autonomous mode would have avoided the accident (or the other way around, with the self-driving system potentially causing an accident). It will record how the car would have acted if the computer was in control, including information about how the car might have avoided an accident (or caused one). That data would then be used to show “a material improvement in the accident rate over manually driven cars,”

With Shadow mode and a huge fleet of cars, Tesla is testing and improving self driving cars at a scale that no other car maker is. Once they have substantial data, Tesla will be able to go to legislators and show them a logbook of all the data and how AP is safer than normal driving. Once a computer can carry out a task more efficiently and more safely then a human, then there will be no reason for it to not get approved.

Thank you. That explains why all all cars are shipped with auto pilot even if the customer hasn't purchased it as an add-on.

I haven't placed a deposit yet but am tempted by one in silver, but I'm also willing to wait for the Model Y if Tesla launch it in 2019. The extra practicality would be useful for long term ownership.

 

[Gianclaudio]

But dear Gian, some have climates where it's freezing the one day and cooking the other

But Arnoud, that's exactly the grace about auto climate controls, it keeps the inner temperature regardless if the outside, warming or cooling as needed. I live in Bolivia's tropical, very hot low lands.....a 2-3 hours trip takes me to the Andean mountains where temp start to drop considerably and yet I don't have to adjust my climate control.

As Cashmere also posted, the higher level of self driving capabilities, the less distracting the use of the touchscreen becomes. We also don't know how good the voice recognition is.

 

[Gianclaudio]

Looks like we have Tesla soldier here If they put dildo on dash, there would be explanation why it is good.

Current Teslas don't feature it, but I heard they hired Gorden Wagoner, so is highly probably for next generation. Pity though, it will use fake metal, chromed plastic balls and purple neon all over the place.

 

[CrunchSlaughtered]

Current Teslas don't feature it, but I heard they hired Gorden Wagoner, so is highly probably for next generation. Pity though, it will use fake metal, chromed plastic balls and purple neon all over the place.

Sensual Purity

 

[EnI]

Surprise! The "Real" Tesla Model 3 Doesn't Cost $35,000

In theory, the Tesla Model 3 is an affordable $35,000 car. But to get a Model 3 worthy of all the hype, buyers will have to shell out a lot more money.

Adam Levine-Weinberg
(TMFGemHunter)
Aug 6, 2017 at 8:15AM

Late last month, Tesla (NASDAQ:TSLA) handed over the first 30 Model 3 sedans to employee-customers at a very public launch event. Now, the electric vehicle pioneer faces the grueling task of ramping production from an average of a little more than 100 per week in the third quarter to an ambitious goal of 5,000 per week by year-end.

Tesla also took this opportunity to offer more details on pricing. As promised, the base version of the Model 3 will have a starting price of $35,000. However, you can't buy that version yet. Furthermore, most customers will probably feel compelled to buy at least some of the available options packages -- which can drive the car's cost way beyond $35,000.

The Model 3 gets a range boost
At the launch event, Tesla revealed that the base version of the Model 3 will offer 220 miles of range, up from a previous estimate of 215 miles. Yet that still lags the 238 miles of range offered by General Motors' (NYSE:GM) competing Chevy Bolt.

THE BASELINE MODEL 3 WILL HAVE 220 MILES OF RANGE. IMAGE SOURCE: TESLA MOTORS.

Of course, Tesla has a big built-in advantage in the form of its Supercharger network. Even fast-charging for the Chevy Bolt is a lot slower than Supercharging a Model 3 -- and those fast-charging stations are few and far between.

Nevertheless, Tesla found another way to outdo GM. A second version of the Model 3 will have a generous 310 miles of range: even more than most of Tesla's pricier vehicles. The extra 90 miles of range represents a big boost in utility, so most potential buyers would be likely to prefer this option. The catch: it costs $9,000 for this upgrade, which is the only Model 3 version that Tesla will build until at least November.

Options are expensive
Other options packages will also add significantly to the Model 3's price. A $5,000 package upgrades the interior to something more in keeping with luxury car standards. It costs another $1,500 if you want 19" wheels instead of the standard 18" wheels, $1,000 for a color other than black, and $5,000 to activate Autopilot. An extra $3,000 will unlock "full self-driving" capabilities -- once regulators give the green light.

Thus, a fully loaded Model 3 would cost nearly $60,000 today. This doesn't even include future upgrades like a dual-motor configuration and "Ludicrous" mode, which will undoubtedly cost even more.


About that tax credit
Tesla appears to be following its well-worn playbook by building the most profitable versions of the Model 3 first, in order to bolster profitability. As long as Tesla has enough orders for the 310 mile range, upgraded-interior Model 3 to fill all of its production capacity, I doubt it will bother building many (or any) of the baseline $35,000 versions.

That's probably rational for Tesla, but it's bad news for anyone looking to save money. Based on Tesla's production plans, the $7,500 federal tax credit for electric car buyers is likely to start phasing out in mid-2018, before disappearing entirely in mid-2019.

Thus, by the time Tesla gets around to producing $35,000 Model 3s in high volume, the full $7,500 credit is unlikely to be available. Anyone ordering a Model 3 today is likely to qualify for a $3,750 credit at best, and potentially less (depending on whether Tesla can meet its production goals).

This isn't a very affordable car
So far, the Tesla Model 3 is getting rave reviews all around. Even electric car enthusiasts who liked the Chevy Bolt agree that General Motors' electric car can't hold a candle to the Model 3.

That's hardly a surprise, though. The Model 3s that reviewers are testing cost upwards of $50,000, and customers ordering today won't qualify for the full $7,500 federal tax credit. This means the final price is well into BMW/Mercedes territory. By contrast, you can walk into a Chevy dealer today and buy a fully loaded Bolt (which is still a far cry from the Model 3) for less than $40,000, while ensuring that you qualify for the $7,500 credit.

Tesla clearly has a legion of fans who are clamoring to get the Model 3 no matter what the price. Nevertheless, I remain skeptical about the long-term size of the market for a vehicle with an average selling price of around $50,000. BMW and Mercedes-Benz each sell fewer than 400,000 vehicles annually in the U.S. across all of their product lines, with no individual model from either one breaking the 100,000 mark in annual unit sales.

For the Model 3 to be the mainstream success that Tesla bulls expect, I believe that the company will need to pack in most of the premium options currently available at a price close to the advertised $35,000 starting price. For now, getting a Model 3 that feels like a "real" Tesla is too expensive for the vast majority of car buyers.

Source: Surprise! The

 

[Deleted member 25]

Lol you have to pay more for any other colour than black. And why do you have to pay a whooping 8000$ for autonomous stuff if all the hardware is already there anyway ?

 

[Sunny]

IMO the full demise of physical buttons is wrong.

Agree with you personally, but yet it happened and if anything, even going to happen further - S8 got rid of even the physical "home" button, making it an "all screen" design and iphone 8 will follow if the rumors are accurate (i7 is already half way there). And this is not "cheap Tesla saving money". Apple and Samsung sells 100s of millions of these devices, the amount of user testing that happens before a decision like this is made is ridiculous - just shows where the user preference is headed. You or I might not agree (I hate even the e-brake being a button vs a lever that deliciously clicks 4-5 times before engaging)

Surprise! The "Real" Tesla Model 3 Doesn't Cost $35,000

Have you configured a 3er or A4 or C class? Model 3 has way more standard features than any of them.

 

[Sunny]

3) Size of Dataset:

I don't know anything about where the all the companies stack when it comes to self driving tech. But I do know a bit about machine learning/deep learning that goes into them. The prefered type of machine learning that is called Supervised learning and it requires lot of "training" data. Example - to teach a machine how to identify a person vs a monkey, you show it 1000 pics of a person, that is labeled as person and 1000 labeled pics of a monkey and when you show it the next pic, the model hopefully can tell the difference. I am oversimplifying things, but you get the essence. More the training data, the better the model. A model build from a million pics will work better than one build using 1000 pics. Hence why size of the dataset is critical for machine learning.

Again, I have no idea, who has better self driving tech, but size of the training dataset is more critical than hardware (or even the software).

 

[Centurion]

Surprise! The "Real" Tesla Model 3 Doesn't Cost $35,000

In theory, the Tesla Model 3 is an affordable $35,000 car. But to get a Model 3 worthy of all the hype, buyers will have to shell out a lot more money.

Adam Levine-Weinberg
(TMFGemHunter)
Aug 6, 2017 at 8:15AM

Late last month, Tesla (NASDAQ:TSLA) handed over the first 30 Model 3 sedans to employee-customers at a very public launch event. Now, the electric vehicle pioneer faces the grueling task of ramping production from an average of a little more than 100 per week in the third quarter to an ambitious goal of 5,000 per week by year-end.

Tesla also took this opportunity to offer more details on pricing. As promised, the base version of the Model 3 will have a starting price of $35,000. However, you can't buy that version yet. Furthermore, most customers will probably feel compelled to buy at least some of the available options packages -- which can drive the car's cost way beyond $35,000.

The Model 3 gets a range boost
At the launch event, Tesla revealed that the base version of the Model 3 will offer 220 miles of range, up from a previous estimate of 215 miles. Yet that still lags the 238 miles of range offered by General Motors' (NYSE:GM) competing Chevy Bolt.

THE BASELINE MODEL 3 WILL HAVE 220 MILES OF RANGE. IMAGE SOURCE: TESLA MOTORS.

Of course, Tesla has a big built-in advantage in the form of its Supercharger network. Even fast-charging for the Chevy Bolt is a lot slower than Supercharging a Model 3 -- and those fast-charging stations are few and far between.

Nevertheless, Tesla found another way to outdo GM. A second version of the Model 3 will have a generous 310 miles of range: even more than most of Tesla's pricier vehicles. The extra 90 miles of range represents a big boost in utility, so most potential buyers would be likely to prefer this option. The catch: it costs $9,000 for this upgrade, which is the only Model 3 version that Tesla will build until at least November.

Options are expensive
Other options packages will also add significantly to the Model 3's price. A $5,000 package upgrades the interior to something more in keeping with luxury car standards. It costs another $1,500 if you want 19" wheels instead of the standard 18" wheels, $1,000 for a color other than black, and $5,000 to activate Autopilot. An extra $3,000 will unlock "full self-driving" capabilities -- once regulators give the green light.

Thus, a fully loaded Model 3 would cost nearly $60,000 today. This doesn't even include future upgrades like a dual-motor configuration and "Ludicrous" mode, which will undoubtedly cost even more.


About that tax credit
Tesla appears to be following its well-worn playbook by building the most profitable versions of the Model 3 first, in order to bolster profitability. As long as Tesla has enough orders for the 310 mile range, upgraded-interior Model 3 to fill all of its production capacity, I doubt it will bother building many (or any) of the baseline $35,000 versions.

That's probably rational for Tesla, but it's bad news for anyone looking to save money. Based on Tesla's production plans, the $7,500 federal tax credit for electric car buyers is likely to start phasing out in mid-2018, before disappearing entirely in mid-2019.

Thus, by the time Tesla gets around to producing $35,000 Model 3s in high volume, the full $7,500 credit is unlikely to be available. Anyone ordering a Model 3 today is likely to qualify for a $3,750 credit at best, and potentially less (depending on whether Tesla can meet its production goals).

This isn't a very affordable car
So far, the Tesla Model 3 is getting rave reviews all around. Even electric car enthusiasts who liked the Chevy Bolt agree that General Motors' electric car can't hold a candle to the Model 3.

That's hardly a surprise, though. The Model 3s that reviewers are testing cost upwards of $50,000, and customers ordering today won't qualify for the full $7,500 federal tax credit. This means the final price is well into BMW/Mercedes territory. By contrast, you can walk into a Chevy dealer today and buy a fully loaded Bolt (which is still a far cry from the Model 3) for less than $40,000, while ensuring that you qualify for the $7,500 credit.

Tesla clearly has a legion of fans who are clamoring to get the Model 3 no matter what the price. Nevertheless, I remain skeptical about the long-term size of the market for a vehicle with an average selling price of around $50,000. BMW and Mercedes-Benz each sell fewer than 400,000 vehicles annually in the U.S. across all of their product lines, with no individual model from either one breaking the 100,000 mark in annual unit sales.

For the Model 3 to be the mainstream success that Tesla bulls expect, I believe that the company will need to pack in most of the premium options currently available at a price close to the advertised $35,000 starting price. For now, getting a Model 3 that feels like a "real" Tesla is too expensive for the vast majority of car buyers.

Source: Surprise! The

The article is utterly stupid and clearly written by a Tesla basher. The $35,000 model exist and will ship this fall. If you don't need 300+ mile range or autonomous capabilities, it will still offer all the tech and amenities you need. As @Sunny mentioned, the Model 3 is well equipped as comes standard with LED head and tail lights, reversing camera, internet connectivity and a large touch screen.

Speccing a 3/C/A4 is a whole lot more expensive and complicated.

 

[DjordjeC63]

Current Teslas don't feature it, but I heard they hired Gorden Wagoner, so is highly probably for next generation. Pity though, it will use fake metal, chromed plastic balls and purple neon all over the place.

At least it will feature something

 

[Cashmere]

Looks like we have Tesla soldier here If they put dildo on dash, there would be explanation why it is good.

Looks like the CEO of Volkswagen is also a Tesla soldier. Being the CEO of the largest carmaker in the world, he obviously knows nothing about this overhyped company. Right?

Volkswagen brand CEO says Tesla has abilities Volkswagen lacks

FRANKFURT (Reuters) - Volkswagen (VOWG_p.DE) the world's largest carmaker by sales, is looking to loss-making startup Tesla (TSLA.O) for inspiration on how to improve its core business, the VW brand's chief executive Herbert Diess said in an interview published on VW's website.

Asked who VW’s main competitors are, Diess told an internal company publication: “In the old world it is Toyota, Hyundai, and the French carmakers. In the new world it is Tesla."

Tesla, headed by Elon Musk, has shaken up the auto industry with its ambition to build a mass market for electric cars, posing a competitive threat to established carmakers, which remain reliant on producing cars with combustion engines.

Diess said the Volkswagen brand, which sold 5,987,800 cars last year, is seeking to catch up with and overtake smaller California-based Tesla, which sold 83,922 cars last year.

"Tesla belongs among the competitors which has abilities that we currently do not have," Diess said in the interview with "Inside", a publication for VW employees.

Around half of Tesla's engineers are software experts, while at VW's core brand it is a much lower proportion, Diess said. Tesla has good electric motors, a fast charging network, autonomous driving technology, internet connectivity, and a new approach toward vehicle distribution.

"This shows that we need to significantly improve. We can do this. We measure ourselves against Tesla quite deliberately. Our goal: Using our abilities not just to catch up, but even to overtake,” Diess, who drives an electric VW Golf, said.

Reuters: Volkswagen brand CEO says Tesla has abilities Volkswagen lacks

 

[DjordjeC63]

^ demagogue.

 

[Centurion]

Looks like the CEO of Volkswagen is also a Tesla soldier. Being the CEO of the largest carmaker in the world, he obviously knows nothing about this overhyped company. Right?

Volkswagen brand CEO says Tesla has abilities Volkswagen lacks

FRANKFURT (Reuters) - Volkswagen (VOWG_p.DE) the world's largest carmaker by sales, is looking to loss-making startup Tesla (TSLA.O) for inspiration on how to improve its core business, the VW brand's chief executive Herbert Diess said in an interview published on VW's website.

Asked who VW’s main competitors are, Diess told an internal company publication: “In the old world it is Toyota, Hyundai, and the French carmakers. In the new world it is Tesla."

Tesla, headed by Elon Musk, has shaken up the auto industry with its ambition to build a mass market for electric cars, posing a competitive threat to established carmakers, which remain reliant on producing cars with combustion engines.

Diess said the Volkswagen brand, which sold 5,987,800 cars last year, is seeking to catch up with and overtake smaller California-based Tesla, which sold 83,922 cars last year.

"Tesla belongs among the competitors which has abilities that we currently do not have," Diess said in the interview with "Inside", a publication for VW employees.

Around half of Tesla's engineers are software experts, while at VW's core brand it is a much lower proportion, Diess said. Tesla has good electric motors, a fast charging network, autonomous driving technology, internet connectivity, and a new approach toward vehicle distribution.

"This shows that we need to significantly improve. We can do this. We measure ourselves against Tesla quite deliberately. Our goal: Using our abilities not just to catch up, but even to overtake,” Diess, who drives an electric VW Golf, said.

Reuters: Volkswagen brand CEO says Tesla has abilities Volkswagen lacks

Hopefully VW iD will be a strong competitor and wont be hampered by other division competing for budget or order up the board/shareholder's priority list.

And why do you have to pay a whooping 8000$ for autonomous stuff if all the hardware is already there anyway ?

1. Simplified manufacturing. The more identical cars are, the cheaper and quicker you can churn them out. Think Ford Model T.

2. Some customers will not want autonomous driving or won't be able to use it where they live so why charge them for it? An added benefit of having the hardware already installed, is that users can pay to unlock the software in the future. Think Playstation Plus or iCloud.

 

[CrunchSlaughtered]

I can't accept the thesis that TESLA can achieve this because it is burning other peoples money and can afford not to earn money(being a start-up), while the established producers are obligated to get certain margins from all their products, which prevents them from trying. BMW created the "i" project exactly with the idea to act like a start-up. They invested more than 2 bln EUR in it and I am sure they burned a lot of money with each car produced ( I don't know even if the turnover from the "i" cars equals the initial investment, not to mention any profit). For me the problem of BMW is that they were surprised by TESLA, and the expectations for the leading role in this field just crashed

 

[Cashmere]

The Model 3 just got its EPA certificate and there's some interesting info

https://www3.epa.gov/otaq/datafiles/CSI-HTSLV00.0L13.PDF

The Model 3 is actually using a Permanent Magnet AC motor (PMAC) compared to the AC induction motors that they've used on the Model S & X. It is quite surprising to see a PMAC motor on the Model 3 as they're known to be more expensive to produce and far more challenging to optimize compared to AC induction motors. On the plus side, PMAC motors have smaller packaging, consume less energy, higher performance and superior longevity compared to AC induction.

In many ways, it appears that the Model 3 is a superior car to the Model S. I suspect we'll be seeing a major revision to the model S before the end of the year.

PS> The EPA lists the Model 3 as naturally aspirated, surely this will appease the purists

 

[Giannis]

(I hate even the e-brake being a button vs a lever that deliciously clicks 4-5 times before engaging)

I hate the e-brake because you can't use it for cornering.

 

[Sunny]

I can't accept the thesis that TESLA can achieve this because it is burning other peoples money and can afford not to earn money(being a start-up), while the established producers are obligated to get certain margins from all their products, which prevents them from trying. BMW created the "i" project exactly with the idea to act like a start-up. They invested more than 2 bln EUR in it and I am sure they burned a lot of money with each car produced ( I don't know even if the turnover from the "i" cars equals the initial investment, not to mention any profit). For me the problem of BMW is that they were surprised by TESLA, and the expectations for the leading role in this field just crashed

I don't buy that thesis either. While there is some truth to it. I think the real resistance and inertia is intrinsic. Again, I will point at the demise of Kodak[1] - they invented digital camera, some of them knew very well it would replace the film and pleaded the leadership to invest in it. Yet the very powerful film division which was responsible for most of profit then, resisted it. When digital camera started gaining popularity, they even spent $500m to develop and market a stupid digital/film "hybrid" version to slow things down, but it was too late by then. History is peppered with these examples.

Same is happening to car industry. There will be many dyed in the wool ICE folk in a traditional car company - up and down the corporate ladder - none of them are going to just roll over and say let us ditch what we did our whole working life and go build EVs instead. Look at just this forum and how many still hold out. Rightly or wrongly, they still will believe ICE has a future and will have to be dragged away kicking and screaming... and this is where leadership comes in. But unfortunately, I don't think BMW has it at the moment. Look at the 'i' brand, which was apparently their electric sub-brand, still someone managed to stick an ICE in each of their models! (Imagine, if i8, instead of the silly 3 cylinder ICE, had a 2nd motor and was the sports car equivalent of a Model S? and i3 had an extended battery version that doubled the range instead of the a generator in the boot?). I know all this will create a virulent reaction from BMW fans, but realize I am one of them, and want to see it not just survive, but thrive.

 

[Busty]

I appreciate your response but you did not provide a single explanation as to why Audi/Daimler/Volvo are ahead. I will try and explain why I think Tesla is leading the self driving race, and i'll summarize it into 4 categories:

1) Developing autonomy on ICE vs EV

Developing a self driving car is very difficult. Developing a self driving car on an ICE platform is VERY DIFFICULT x 100 - even more lines of code need to be written to communicate the self driving software with the transmission, air/fuel flow, and the thousand other parts that go into an engine. Now, if that is not complicated enough, you can add an extra layer of complexity by introducing hybrids (ICE + battery + electric motor). Its an even BIGGER cluster-fu*k. Making hybrids in the early 2000's made sense to familiarize car makers with how batteries and electric motors work together while they gradually phase out ICE. Making hybrids in this decade is regressive. They add an extra layer of complexity to everything from manufacturing cost/process to repairs - solely for the reason of showing pseudo emission numbers. They don't move the industry forward - they just tick a box in a sandbox test that says "when X car is driven at Y speed it emits less than Z emissions".

Add to that, the current fuel station infrastructure is not designed for automatic refuelling. A human is needed to pump the fuel into the car. So for cross country travel, all petrol stations would have to be upgraded. This is another anomaly beyond the control of ICE carmakers as they would have to beg their brethren at the OPEC Cartel to make upgrades.

These are obstacles that Tesla does not face. The simplicity of EV cars (far fewer parts) means that Tesla's pace of software development is leaner and significantly faster. Tesla is also vertically integrated with its own supercharging network. If they decide to incorporate wireless charging at all their stations tomorrow, they can easily do so without needing to wait for anyone.

2) Democratizing Technology

Chapter 1, Page 1 out of every successful tech startup's guide to success is making your technology scalable. This usually entails suffering losses in the beginning but once you scale upwards, thats when you begin to reap the benefits. Last year September when Tesla showed off Autopilot 2.0 with the 8 cameras they showed the following video:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

If you watch the video, you can observe how camera is capable of discerning between lane lines, in-path objects, out of path objects, road signs, road lights and motion. The amount of data 1 camera captures is staggering. With 8 x cameras + radar & ultrasonic sensors you are crunching through teraflops of data. To crunch through that kind of data you need a supercomputer.

Having these array of sensors paired with a super computer on a $70,000 car is impressive. Having it on a on a $35,000 car is disruptive. Do you think the Germans will ship this kind of hardware on a base C Class, A4, 3 series? Highly unlikely because they're set in their old ways of making high profit margins and reserving cutting edge tech for the S Class/A8/7 Series and only updating their cars once every 5-10 years. Even if they miraculously decide to ship this type of hardware on all their cars from next year, do you think they would be able to sell it for $35K and still make a profit? The BMW 320i costs $34K and they still make you pay extra for a reverse camera.

3) Size of Dataset:

You may have heard, data is the new oil?

If you were to take all the self driving cars from Google, Mercedes Benz, Volvo, BMW, Audi, Volkswagen and add them together, you would find that they probably have less than a thousand test cars running around in their self driving fleets. To be generous, lets just assume they have 5,000 cars.

So far in 2017, Tesla has already sold +50,000 cars with all of the above hardware. Tesla already has a fleet of test cars that is 10 fold bigger than all of their rivals put together! By the end of 2018 (if the are able to deliver Model 3s), there will be +500,000. The company with the biggest dataset will also have the most refined algorithm for self driving cars. With every car Tesla sells, they're putting an exponential gap between them and every other carmaker that doesn't sell a car with self driving hardware.

4) Legislation and shadow mode

When Elon Musk introduced AP2 last year, he spoke about Shadow mode.

Before Teslas can start driving autonomously, the company needs to collect a lot of data to prove to customers (and regulators) that the technology is safe and reliable. So, the car will run Autopilot in “shadow mode” in order for Tesla to gather statistical data to show false positives and false negatives of the software. In shadow mode, the car isn’t taking any action, but it registers when it would have taken action. Then, if the Tesla is in an accident, the company can see if the autonomous mode would have avoided the accident (or the other way around, with the self-driving system potentially causing an accident). It will record how the car would have acted if the computer was in control, including information about how the car might have avoided an accident (or caused one). That data would then be used to show “a material improvement in the accident rate over manually driven cars,”

With Shadow mode and a huge fleet of cars, Tesla is testing and improving self driving cars at a scale that no other car maker is. Once they have substantial data, Tesla will be able to go to legislators and show them a logbook of all the data and how AP is safer than normal driving. Once a computer can carry out a task more efficiently and more safely then a human, then there will be no reason for it to not get approved.

The adoption of self driving cars will happen far quicker then many people expect. While the rest of the car industry is trying to figure out how to make affordable electric cars, Tesla's secret weapon is actually its lead on self driving capabilities. IMO, Model 3 is arguably the most important car to come around since the birth of the automobile.

Thanks a lot for your elaborate response. Nevertheless, it sounds like a rather uncritical adulation by a EM/Tesla fanboy tbh. Also, some of your 'facts' are highly disputable.

"you did not provide a single explanation as to why Audi/Daimler/Volvo are ahead": yes, I did. I used a concrete autonomous function in order to illustrate that i) hardware is not everything that defines autonomous functioning in automotive use cases and that ii) Tesla does pretty little real-life testing during development and leaves quite a lot of those efforts (and some associated risks which are handled via disclaimers) to their customers. As long as the majority of cars on the street are not autonomous, I (and quite a few specialists in the field) consider that solution sub-optimal. At least, it should not be seen as the golden standard (which you do when people say 'they are lightyears ahead').

@1) We are addressing different paradigms. You are mostly talking about BEVs and IT which is legitimate but a different story. I explicitly excluded those topics as my experience says that the vast majority of customers still see cars as rather mechanical products. In fact, one reason the big OEMs don't take their cars more 'online' is that they know many dislike the idea. But this does not mean extensive research takes place.

@2) Pseudoargument imo. I call it democratizing when they sell a million cars with all of that equipment and really have to earn money with their products. Without investor pressure etc., all other brands would also make gifts and 'democratize' tech. Btw, the sensor tech, at least in Model S and X, is pretty much standard as far as the supplier components are concerned.

@3) 4) More or less one argument (the what? and the how?) and oh yeah, there we go, another shibboleth: data is gold. Of course, every child knows that today (like the 11th commandment) but, at the same time, YOU may have heard: big data alone starts to get out of fashion again. Only lately, great Science article on it. Fact is: big data alone teaches nothing as long as it's not processed, analyzed, and interpreted correctly. 'Correct analysis' not only refers to hard quantitative data (such as false positives or false alarms and other within-system variables) and resulting correlations etc., but also to semi-quantitative or even qualitative data such human behavior/error/cognition, as well. In the test protocols I know or heard about, Tesla focussed practically only on the former but ignored the latter.
Also, datasets of the big OEMs are much bigger than you think. And I don't only refer to the Mini or E-Tron fleets but to various series vehicles that have been data logged in the last decade.
Information is mostly readout during service. That is why the OEMs want and need their customers in their authorized repair shops: it is the only place to make money (selling cars is not the most important source of revenue anymore) and it is where all the data comes together. As mentioned above, permanent data collection would not be accepted by most customers.
And here we are again, with our different premises: you test the others by Tesla standards, and I compare Tesla to industry standards. At the moment, I consider that much more appropriate. Of course, the industry has to move (a lot and quickly). But not necessarily in Tesla's direction, imo. To claim that the "Model 3 is arguably the most important car to come around since the birth of the automobile" is a very bold statement, mostly because you argue (as EM and Tesla itself) primarily in hypotheses and predictions. Bad thing is that in the past they did not even predict their sales numbers correctly. And: > 35 Grand are still very much money for very many customers. Imo, the Model 3 is another (smaller) luxury toy, as well.

However, PMAC may indeed be a major improvement over most current BEVs, I have to admit. But is there really nobody who knows how much this baby (the whole car I mean) is weighing? This info is crucial for the full picture I guess.

Edit: Now, I heard weight for standard range is 16xx kg, long range 17xx kg. That is not as bad as I had expected.

 

[EnI]

I don't buy that thesis either. While there is some truth to it. I think the real resistance and inertia is intrinsic. Again, I will point at the demise of Kodak[1] - they invented digital camera, some of them knew very well it would replace the film and pleaded the leadership to invest in it. Yet the very powerful film division which was responsible for most of profit then, resisted it. When digital camera started gaining popularity, they even spent $500m to develop and market a stupid digital/film "hybrid" version to slow things down, but it was too late by then. History is peppered with these examples.

Same is happening to car industry. There will be many dyed in the wool ICE folk in a traditional car company - up and down the corporate ladder - none of them are going to just roll over and say let us ditch what we did our whole working life and go build EVs instead. Look at just this forum and how many still hold out. Rightly or wrongly, they still will believe ICE has a future and will have to be dragged away kicking and screaming... and this is where leadership comes in. But unfortunately, I don't think BMW has it at the moment. Look at the 'i' brand, which was apparently their electric sub-brand, still someone managed to stick an ICE in each of their models! (Imagine, if i8, instead of the silly 3 cylinder ICE, had a 2nd motor and was the sports car equivalent of a Model S? and i3 had an extended battery version that doubled the range instead of the a generator in the boot?). I know all this will create a virulent reaction from BMW fans, but realize I am one of them, and want to see it not just survive, but thrive.

Oh, c'mon. The devil is in details. Comparing very complex automotive - (eg a vehicle has to comply to very strict safety, emission etc standards; it has to last longer; it has high price and it's not instantly replaceable like any other consumer products like phones, cameras etc) - to consumer (electronic) devices is quite ignorant.

Also ... Kodak did actually bankrupt but eg. another very big & strong film maker - Fujifilm - didn't. And even mobile phone giants like phone giants Nokia, Ericsson & Motorola did fail to offer smartphones, other "dumb"phone makers like Samsung, LG, Sony didn't - and they are offering quite good and capable smartphones indeed. Etc.

Also car development & production is a bit more complicated then development & production of consumer devices. Therefore the shifts in automotive industry are much slower.

Nobody does deny Tesla is currently leading the play ... But to say others will be obliterated like Kodak or Nokia, is very much absurd!

Not to say German carmakers - especially BMW (although I don't know why exactly) - have been under fire (I guess due to Dieselgate & cartel allegations) despite they pour billions of Euros into BEV and Autonomous driving development. Much more than perhaps other carmakers. Nobody is talking about GM, Ford, FCA, Toyota, HyundaiKIA, Renault Nissan, PSA, Tata and bunch of Chinese carmakers ... or even smaller ones like Mazda, Mitsubishi, Suzuki, Honda, Jaguar Land Rover etc.

Regarding BMW i ... it has never been a sub-brand for electric cars --- geez, even i8 is a HYBRID --- but a sub-brand for introducing new tech. CFRP body frame & body panelling was introduced in series production car; a performance hybrid was introduced; new natural & recycled materials were used in the cabin; new ways of design were shown (layering - due to CFRP); first series EV was introduced; REx tech was introduced It's a niche sub-brand for techy cars. Therefore iNext will introduce some other new technology. Just like BMW M stands for Motorsport (sporty BMWs) the BMW i stands for Innovation. Not for BEVs necessarily. The problem is the cars & tech were more experimental & engineering-driven and unfortunately not very customer-driven. That's why some solutions are quite dorky.
When it comes to electrification ... BMW will offer BEV variants of their core models, along with PHEVs & MHEVs.

Current BEV hysteria is just a hysteria. Sure cars will transition to BEVs. But not over night. Even Tesla is limited by charging stations (more cars will they sell, more charging stations will be needed); by production capacities; by product price (subsidies) etc.

And not many have a house in suburbia with charging possibility @ home. Many around the world live in apartment blocks or old(er) city buildings with no yard, with no (at least not enough) public charging points etc.

BEVs need mass volume charging infrastructure (in the streets; @ rest stops; @ at shopping centers etc). And it seems electricity providers still aren't really interested in building the charging networks. So, before a BEV boom can happen the charging issue has to be resolved.

Mind in 2016 over 72 million passenger cars were produced worldwide (more than 24 millions in China alone). Only 1,2 million of them were BEVs. Roughly 1.7% only. And even when Tesla does produce 1 million BEVs per year - that's still a peanut compared the whole global car production pie. So, to say the "old" car industry will go down due to Tesla is a moronic claim. Tesla is & will be a niche player - offering premium hi-tech cars in rather limited volume. Other carmakers will transition to BEVs slowly & continuously ... waiting for battery plants to be built, charging network to be built etc. Sure some smaller local markets (like eg. Norway) will transition to BEVs much quicker - and there carmakers with BEVs in portfolio will replace the ones without BEVs. But largest markets will transition to BEVs in decades ... so there will be plenty of time for ANY carmaker to adapt & to transition.

And just like with ICEVs ... Not evey BEV will be the best. Some will be better than others. Perhaps Tesla will offer the best BEVs & best AD cars ... But their production will be limited & their brand image will dictate premium prices ... Others will have to buy a Ford BEV or VE BEV or Hyundai BEV etc ... with less performance, with less millage, with slower charging etc.

Again ... please don't compare consumer device sector to automotive sector. It's completely different story; with different rules & different dynamics. We don't compare aviation sector with automotive sector either, don't we?