Tesla will not be able to scale its manufacturing capacity | Hacker News Reader

Tesla will not be able to scale its manufacturing capacity

bloomberg.com

79 pointsOsiris30121 comments

First it was "electric cars are not viable technologically, they have no range, there is no charging infrastructure, etc."

Then it was "electric cars are too expensive, they are rich peoples' toys, and mainstream consumers will never want them anyway."

Now it is "Tesla will not be able to scale up to meet all the demand."

The fact that EV naysayers have been forced to cycle through this spectrum of responses in less than 10 years should be a clue of some kind.

Teslas are still "rich peoples' toys" some places, model s cost 120000 USD for the cheapest model.

Tesla's are really affordable cars in other places (Norway), due to tax incentives for electric vehicles over vehicles.

Not really, the basic no frills model S starts at over $80k. Sure that's a lot cheaper than other cars in the same performance class, but that's a far cry from being affordable.

A car that still costs 60.000EUR (and that's a really stripped down model) after incentives is nowhere near "affordable" for most of Europe, much less world.

There is a big difference between Electric Vehicles as a whole and Tesla. Electric Vehicles are 'probably' here to stay and will 'probably' eventually make up a significant portion of first world mainstream automobile market. Whether Tesla significantly leads the industry in volume of Electric Vehicles manufactured and sold is another matter entirely.

The main component one has to not underestimate is that in the automobile industry the technology/engineering problem of the car is less important to the technology/engineering/process of the plant itself.

Toyota is not a lumbering clumsy inefficient organization, it is a sharp nimble efficient organization... It only makes rational sense that Tesla should follow their lead

>>The main component one has to not underestimate is that in the automobile industry the technology/engineering problem of the car is less important to the technology/engineering/process of the plant itself.

What you say makes some sense, though, you seem to forget that there is vast difference between the technology/engineering/processes of the plants needed to manufacture the electric cars and the ones needed to manufacture the fossil fuel cars.

Another thing, it should not be too difficult for us to grant that Elon Musk is not a fool to make that big a statement without doing some homework. Also, don't forget the current state of the art of the technology/engineering/processes he may employ now that Toyota could (and still can) only dream of.

What are you talking about. I mean specifically. What state of the art manufacturing/engineering/processes do you specifically speak of that Toyota can only dream of that Tesla can employ?

Are you talking from facts? OR are you talking from feelings, and hunches, and faith in Musk?

Space X, is one thing... in that bracket Musk is competing with inefficient opponents who have historically benefitted from lack of competition, crony government contracts, in an industry that is primarily product engineering driven that requires a small volume manufacturing of extreme high complexity vehicles

In Tesla he is competing against highly efficient opponents who have been in a dog eat dog survival of the fittest competition for going on 40 years which is only somewhat product engineering driven and high volume manufacturing of relatively low complexity vehicles.

Tesla can overcome this... possibly but it is a huge challenge that a responsible investor cannot sweep under the rug...

What are =you= talking about? Tesla's manufacturing capabilities are far more modern and informed by Silicon Valley. Electronic cars (outside the Model X) have far fewer parts and are substantially less complex than combustion engine. Tesla has =far= better software development capabilities which is becoming much more important. Tesla has a much simpler distribution network and much closer contact with users. etc, etc, etc. It's hard to make much of a comparison.

Then why did Edmund's car need FOUR drive units in two years?

http://www.edmunds.com/tesla/model-s/2013/long-term-road-tes...

That's the FOURTH drive unit (3rd replacement) their Model S needed.

The fact of the matter is, my Ford Focus (and most people's Toyota Camry from the 90s) has needed less maintenance than than the two-year test Edumnds did on the Model S.

Unless you like paying $90,000+ for a car that dies in the middle of the highway as the drive unit completely craps out, I'm not sure if the Tesla Model S is actually a sign of Tesla's manufacturing prowness.

http://www.edmunds.com/tesla/model-s/2013/long-term-road-tes...

And considering the widespread reports of issues with the "simple electric" Drive Unit, I'm thinking the electric drive unit is a lot more complex than Tesla's marketing slides indicate.

---------

Worst part is that this isn't anecdotal evidence. MULTIPLE reporters have had their Tesla Model S die on them in the middle of the highway.

https://www.youtube.com/watch?v=0eTeJMACrtM

That actually is anecdotal evidence.

Anyway, the Model S was explicitly the stepping stone project to get to the Model 3 (i.e. Musk stated this). They've probably learned a lot.

> That actually is anecdotal evidence.

If it happened ONCE its anecdotal evidence.

If it happens repeatedly to different reporters around the world? That's a statistic. The Model S drive unit is not reliable. Period. It doesn't matter how "simple" an electric motor theoretically is, in practice, the Model S motor has clear drive unit issues.

> Anyway, the Model S was explicitly the stepping stone project to get to the Model 3 (i.e. Musk stated this). They've probably learned a lot.

I thought the "Roadster" was when they were supposed to get all of these issues figured out. The Model 3 has to be built in a year in capacities of (wtf?) 500,000 / year by 2018 or something.

There's no room for experimentation here. Elon Musk is ramping up production extremely fast and has set stunningly high expectations.

---------

Now, don't get me wrong, I think Tesla is a great company with a great vision. And that Musk is doing his best. But what I'm trying to say is that this company's balance sheet, and reliability record with their cars is causing me some concern at this point.

I would like it if Tesla could succeed, but I'm not going to hold back on criticism just because I like their mission.

No, it's still anecdotal evidence unless you actually have a meaningful statistic gathered through scientific inquiry. If you have just heard several people complain about failures, that's anecdotal evidence. After all, that is precisely the definition of "anecdotal evidence."

You really don't know, do you?

http://www.consumerreports.org/cars/tesla-reliability-doesnt...

> As part of our Annual Auto Reliability Survey, we received about 1,400 survey responses from Model S owners who chronicled an array of detailed and complicated maladies.

I'm not kidding when I'm saying this "isn't anecdotal" anymore. Over a THOUSAND reports have come in, and the consensus is that Tesla is NOT reliable.

Tesla reliability not unlike other performance vehicles that are version 3, 4 or 20. That tesla has achieved what it has on v1 is amazing.

Consumer Reports: "The Tesla wasn’t the only high-performance vehicle that fell below average in reliability. Others include the BMW X5 and 5 Series, and the Chevrolet Corvette."

Yes. Lets compare Tesla with the worst cars on the market in terms of reliability. And then demonstrate that other cars can have problems too.

Or you know, as a smart consumer, I'll be looking at a workhorse car like a Honda Accord, Ford Focus, Toyota Camry, etc. etc. and not have to worry about reliability issues.

Or if I'm actually going to buy on Luxury + reliability: the highly reliable Lexus ES looks like a good pick at half the price of a Tesla.

I'd say the German luxury brands are the most appropriate comparisons.

Sure. But you don't get to pick the least reliable model of the NINE lines of SEDANS that BMW makes, and then decide that its a fair comparison.

Tesla has two lines of cars right now, and only a third one planned. Their entire focus is supposed to be on the reliability and features of these two lines they're making.

Tesla's are generally reliably. I bought an original one in 2012, then I upgraded to AWD when that came out. Never had a problem other than "slight amount of moisture in my tail light" which they fixed. There have been a number of reporters who appear to have an anti-tesla bias, notable the NY Times report Broder. His name has become an adjective ("That guy brodered you"). They are physical objects, they can fail of course, but generally they are very reliable, and in my experience owning them over 4 years, have been without issue.

On the other hand, THIS is anecdotal evidence.

Look man, I have a Seagate 3TB hard drive. It hasn't failed on me and I bought it damn cheap. But when the statistics come in...

https://www.backblaze.com/blog/wp-content/uploads/2014/09/bl...

I'm not going to argue against them. Consumer Reports (survey methodology) and Edmund's (single case study) have demonstrated the issue. Tesla cars are NOT reliable, compared to Toyota Camry or Corollas or whatever else is normal out there.

Does that include the episode on Top Gear that was scripted into breaking their Tesla down in the middle of a test run?

According to your logic, 3 percent of scientists deny global warming, so that is a statistic we should take seriously as well. I don't care what denialists say because I can appreciate the fundamental science behind global warming.

You seem to lack the vision to see that simplicity in design always wins out. It does so in nature, and it will do so in our automobile engines long-term.

You are suffering from an extreme case of cognitive dissonance my friend.

In a world of 7 Billion people, every anecdote has become a statistic. Figure it out or get left behind.

> Does that include the episode on Top Gear that was scripted into breaking their Tesla down in the middle of a test run?

Edmunds is a serious car site, and not an entertainment show like Top Gear.

The statistics include Consumer Reports, who surveyed over 1400 Tesla S owners and have come to the same reliability issue as Edmunds. http://www.consumerreports.org/cars/tesla-reliability-doesnt...

---

And unlike Top Gear, which seems to be an anti-Tesla idiotic entertainment show. Consumer Reports was pro-Tesla until the survey came in.

http://www.cnbc.com/2015/08/27/teslas-p85d-is-the-best-car-c...

Unfortunately, Consumer Reports does not do long-term reliability tests (see Edmunds for that sort of stuff). Instead, Consumer Reports has a very robust survey methodology. They rely upon a survey conducted at a later date. So when their survey results came in that documented the large number of Tesla problems, they had to retreat from their superb rating.

A 103 out of 100 by Consumer Reports (initially). And even then, the Consumer Reports guys could not refute the reliability concerns.

+1 for sticking to the engineering arguments and ignoring the ad hominem.

Part of it is acknowledging that the Top Gear show was truly awful though. It was unfair for sure, and I can see why someone who was only familiar with the blatant anti-Tesla Top Gear episode would be distrustful of these sorts of arguments.

Yes, there was an anti-Tesla conspiracy going on a few years ago. But I don't think we're at that phase anymore. Certainly not with these Consumer Reports surveys or Edmund's tests, which seem to check out as legitimate.

In the early days of Tesla Model S, they strived to get people back on the road quickly, and they replaced entire subsystems when they didn't really need to. If you search the news group "tesla motors club", you'll find discussion of how telsa found a number of tiny tweaks that they could use to adjust some of these minor issues. They were not drive train failures really. I have driven about 40k miles, no drive train failures (anecodotal, blah blah).

Even if they weren't technically drive unit failures, it is clear that a number of Tesla Model S cars have failed in some form. They were forced to be taken back to the shop and "fixed".

The reports are clear: Tesla has excellent consumer service and warranties. Owner satisfaction is EXTREMELY high, despite the issues. So Tesla is doing everything they can to address the issue, and this is commendable.

However, if you actually compare the reliability scores on pure numbers, it is clear that the Tesla Model S has a problem. It may not necessarily be drive unit related, but there is clearly a problem that affects a huge swath of people.

Tesla has one facility, which is definitely state of the art, and has some of the largest automobile manufacturing robots in the world... on the other hand toyota has about 40-70 plants of varying level of technological advancement... its quite a range because some are only operated by toyota not owned... there more recently opened factories such is the one in Mississippi are pretty much universally accepted amongst the most advanced in the world... In fact Tesla's current factory is actually an upgraded old ford/toyota partnership which was upgraded as part of a tesla/toyota partnership...

which is all to really say... 1 plant vs 40-70 plants... 90K vehicles produced per year to 9,000K vehicle/year... unproven manufacturing prowess vs universally accepted best manufacturing capability

toyota is universally accepted as having the most advanced automotive manufacturing as far as major manufacturing... Even Elon Musk wouldnt have the balls to make the claims you guys are making here... please lets stick to reality.

i wish tesla well... they i think are going to be the main catalyst in popularizing the electric car... will they actually hit 500K units in 2018 out of the Fremont factory? seems unlikely given their track record but maybe... i mean toyota was able to produce that much out of the same facility before it was upgrade about a decade ago so its possible...

Companies always start small.

Your argument is basically "the small company is small". So what? Do you propose that business can work in some other way? How else would you propose that it works?

Elon setting aggressive goals is a management tactic and it works well for him. Sure, it'll probably slip till 2019. But if the goal was 2019, it would slip to 2020.

> there is vast difference between the technology/engineering/processes of the plants needed to manufacture the electric cars and the ones needed to manufacture the fossil fuel cars

Do you have any good articles that explain this difference? I'm honestly curious what changes when you change the propulsion system.

It's also very likely that the car makers of the future will be the ones that are best at software, not hardware, ironically enough.

Tesla, while an engineering wonder, is much less complex to produce than cumbustion engine vehicles. It's a wheeled computer.

I suspect their challenge will be more about getting enough resources for batteries and finishing the production plant.

Remember, right after apple launched the iphone and experts claimed smartphones were for the first world while growth would be in the third world?

Having issues with, say, your gull-wing doors is not specific to electric cars. Same with the issues they're having with the Model 3 and suppliers not being able to meet the dates they want. (Drive units dying is specific to electric cars, but that's also likely to just be teething pains.) There's a lot more to a car these days than a power plant.

The engine is, comparatively speaking, the solved problem for car reliability these days, at least in the first 100k or 200k miles. One of the most common characteristics of unreliable or fancy traditional cars is flaky electronic systems, as Telsa is as complicated as anyone there.

Those statements are all still largely true, they haven't been "cycled through." There are a lot of good things happening in the world of electric cars, but right now, for most people, they're too expensive and they require sacrifices in utility and convenience that aren't seen in ICE vehicles.

The Model X has 250 mile range. But you think it's "largely true" that electric cars have a range problem?

There are ~600 charging stations worldwide, and that's increasing to 1200 during 2016. But you think it's "largely true" that there's no charging infrastructure?

Not the person you're responding to, but yes and yes.

600 charging stations worldwide? That's worse than I thought. I'm traveling across country as we speak (I live in an RV). I've seen one electric car charging option in the past few weeks...I've filled up over a dozen times in that period. Realistically speaking, you cannot take an electric car on vacation.

That doesn't mean I don't love electric cars (I even held TSLA for a while before needing to sell some stocks for a major purchase...I kept my GOOG and sold my TSLA).

One can love electric cars while still acknowledging what they cannot yet do. There was a time when horses were a better cross country means of transit than automobiles because there wasn't infrastructure for cars yet. We're at that stage for electric cars. It'll pass.

Maybe that's 600 Tesla provided charging stations - there ~4000 electric car charging stations in the UK alone:

https://www.zap-map.com/statistics/

You may not have seen that many dedicated electric car charging outlets, but I bet many of the RV parks you've stayed at have a 220 V / 50 A NEMA 10-5 connector, which can give you about 30 miles of charge per hour in a Tesla. Not a supercharger, but decent.

Availability of non-Tesla electric charging depends vastly on where you are. In the northeast, there are a bunch of places that have 2-4 EV spaces with a CHAdeMO connector which (anecdotally) are always full. So if you're planning a trip cross-country, you basically have to stick to the interstates for superchargers.

If you have an RV, you probably think that traveling cross-country via only the interstates is a pretty bad way to see the sights though :).

600 Tesla stations alone. Don't forget you can charge your car from the regular outlet, of which there are billions.

Spoken like a true believer. 600 charging stations worldwide, eh? There are probably 600 gasoline stations in my metro area.

Therefore, emphatically Yes, it's "largely true" that there is no charging infrastructure.

All of the things people have cited as challenges for EVs are still true:

- limited charging infrastructure - limited range - limited data on longevity (look at Leaf's battery life)

The range issue may be moot for commuters, who drive 40-50 miles. They can also plugin at home, and recharge at night.

EVs will grow in usage. Slowly.

The question of whether Tesla can scale up its manufacturing operation is a different one. Unrelated to EVs.

There are 600 "super chargers". They are 100k+ regular high powered chargers. At my office alone there are 20 charging stations, including j1772 and high powered (40 amp) outlets.

EVs have a range problem. 250 miles is fine for day to day driving, but most Americans routinely need to travel farther than that. Travel to nearby major cities, visiting relatives, vacations, etc. The supercharger network is a long way from being about to support travel for most people, and even once it is the charge time (30 minutes for 170 miles range) is a significant inconvenience on any trip long enough to require recharging.

The bottom line is that even if an EV meets your daily needs, it isn't going to be good enough to be your only car. That really isn't going to be the case until EVs regularly have a 350-400 mile range, the supercharger network is reasonably dense, and a supercharger can top off your car in 10-15 minutes, tops.

I love electric cars but I totally agree that it's largely true that _most_ electric cars on the market have a range problem (eg cars that are priced in a way that they are _almost_ accessible for the general population).

They do have a range problem, but not in the way that you probably think.

If you take a look at the affordable EV segment, in the 2011 to 2016 era, it has been represented only by Nissan Leaf.

People are reporting anywhere between 15-30% of capacity loss within the first 45,000 miles. The problem with battery powered vehicles is that there's a guaranteed loss of capacity after N charge cycles, which currently seems to be ~600-800 charges (77 miles of range x 600 charges = 46,200 miles)

Teslas have obviously not run into this problem yet, because their range is 200+, so you'd only start seeing significant capacity losses at 120,000 miles driven, which I'm guessing not many people have done since 2012.

But for smaller range vehicles this presents a significant problem: the commute you are able to make in the beginning of ownership may suddenly become impossible in just 3-4 years of driving. And how do you get more range? You need more batteries.

In fact the first vehicle to solve this and have a decent enough capacity buffer at an affordable price will be Tesla 3.

On the other hand, this will put the nail in the coffin of the "not being able to scale" argument. In fact, Tesla 3 will be successful ONLY if they're able to scale.

Yep. I have just one car, and I don't want to rent a car every time I need to drive longer than half the day. And I certainly don't want to turn a long trip into a day-long or days-long trip because I need to stop and recharge my battery every so often.

Electric cars are getting there, but you need to be in a specific location and have specific requirements for owning one to make sense.

"Shipping is a feature." Imagine someone rear-ends your current gas-guzzler and the insurance company totals it. Now try and buy one of the damn things. Over the air suspension upgrades are great. So is being able to walk onto a lot and drive a car home.

EDIT: "Pretty sure you can have one delivered like, next day." Love the HN optimism. As if my comment didn't imply that I just got fuckin' rear ended and tried to buy one of the damn things. Website says June on the S. I was told July. X is several months out (no firm date).

You know what really sucks about talking to a salesman on the lot? Being told, "ain't got any and you can't have one."

EDIT 2: CPO? If I'm going to pay $130,000 for a used car it better come with C3PO. The pre-owned market is warped because you can't buy a new one. It's all part of being unable to meet demand. My argument stands.

Pretty sure you can have one delivered like, next day. And not have to deal with the salesman on the lot.

Tesla has CPOs and inventory cars you can get much faster. Of course you're limited in your choices then, just like buying inventory from a dealer.

Most of those things are still true. A duded-up Honda Accord is a better value than an Everyman Tesla in every category.

How about safety?

This is just crazy wrong. The Tesla S is insanely more safe, energy efficient, gets about the same amount of range, and will absolutely smoke an accord in a 0-60 and 1/4 mile race. Are you saying its a better value because it would be less expensive? You could say a $20 POS android phone is a better value than a galaxy s7 because it costs less and does mostly the same stuff, but it doesn't make that true.

> and will absolutely smoke an accord in a 0-60 and 1/4 mile race

Yeah, that's just the right argument. I'm sure you are talking about taking your car to the track, no driving like a reckless idiot.

Funny thing is, Tesla Model S isn't your best value for money as a track car either.

Its 4500lbs minimum, the P85D is even heavier (larger batteries I guess?). Its got a skidpad turn of .85g, while the good high-performance cars are .95g or higher.

A $25,000 Subaru WRX achieves 5-star safety rating and gets 1:45 Laguna Seca lap time.

The Tesla Model S "Performance Edition" gets a 1:48 Laguna Seca lap time.

https://forums.teslamotors.com/forum/forums/model-s-laguna-s...

I think its weight forces the Tesla S to take slower and wider turns, while the more nimble cars manage to beat it on the Track.

Once you leave the 30k price range and into the cars that are ACTUALLY competing against the Tesla Model S, you start getting to Cadelac CTS-V and faster cars.

https://www.youtube.com/watch?v=la6BZ89Z6bs

1:38 on the CTS-V.

And yes, the CTS-V is a large 4000lb luxury 4-door sedan. Its price, size, and weight-competitive against the Model S.

Heavier cars also wear through tires faster and put more stress and heat on the tires/brake pads/rotors.

Speaking of heat, I hear that Tesla's motors and brakes heat up a lot more on a track. To the point where the brakes and engine start throttling the car down a bit after a lap or two.

Which is fine. My personal car of choice apparently overheats its brakes on the Laguna Seca under stock settings (Ford Focus, although the recent "RS" edition looks to have significantly better specs for a track). Its a challenging track. So the Tesla Model S isn't the only car that is unable to sustain high-performance throughout an entire course.

But lets not pretend that the Model S is a real performance car or anything, not until it starts placing somewhere impressive on a track somewhere (Laguna Seca, Nürburgring, or whatever)

--------------

http://jalopnik.com/heres-what-a-tesla-model-s-can-do-around...

> "The lap itself was around 10 minutes Bridge to Gantry (in heavy traffic) but unfortunately the car went into a reduced power mode about 3 minutes in due to excess battery heat (at least, that's my guess)."

So there's probably some heat issue with regards to Tesla Model S on serious test tracks.

All in all, the Tesla isn't a "performance" car, although it "isn't slow" by any stretch of the imagination.

I suspect "better value" in this case means the total cost of the car over its lifetime. Given the reliability problems of the Tesla Model S and the high sticker price, I would bet that the Honda Accord has a much better lifetime value. Maybe when Tesla fixes their manufacturing problems and releases a Model Average, that will change.

Like 99.8% of drivers, I don't race my car, so that stuff doesn't really matter. If you care about that, compare to a BMW 5 series.

I used an Accord as an example because I'm familiar with it, feel free to pick almost any midsize sedan to reach the same conclusion. Perhaps a Volvo or Subaru would address whatever your safety concern is.

If you look at the 90th percentile car buyer driving 12-15k/year, there is still a total cost premium from an operations POV. Insurance will skew that more.

In terms of range, Tesla makes a compelling case for a typical commute. But any distance driving is a different story. Charging on the go is both sparsely available and very time consuming.

I'm not saying it's a bad car, just a poor value for most people.

No offense intended but you come off sounding like a True Believer here. There have been and continue to be severe limitations on electric cars:

- high cost of batteries

- limited range largely limits them to urban settings

- materials currently required for manufacturer are limited

- cold weather reduces battery effectiveness and thus range

- time to charge is a huge issue compared to simply refilling a tank

- charging stations require a massive and expensive infrastructure upgrade for widespread use

- cost

- unproven reliability. This is partly an issue of the technology being so new but even Tesla says the batteries need to be replaced after ~8 years.

Battery technology has come a long way in the last 20 years. It's already enabled a bunch of applications that are hugely significant (eg smartphones, drones) to the point where I think that a massive breakthrough in battery technology may well be the most important discovery of the 21st century as the consequences will be so staggeringly massive.

So I'm not saying the electric vehicles are the future or that they aren't. The future is hard to predict. Regardless, putting liquid fuels into vehicles is here for some time to come yet.

Time to charge is not a huge issue, you can simply replace the battery for a full one. Tesla has an automated battery swap station that can replace one in 90 seconds.

Those battery swap stations were mixed (... For now?)

http://www.roadandtrack.com/new-cars/car-technology/news/a25...

I think you mean "nixed".

Whatever the case, I tend to agree that battery swaps is not a scalable solution.

Agree with whom? Elon said he's not pursuing them because they are not popular, as Superchargers are fast enough. That's in direct contradiction with your claim that time to charge is a huge issue.

> you can simply replace the battery for a full one.

I don't think so. The Tesla Model S batteries are everywhere through the chassis. There is no single "battery" to replace. It would be major surgery to swap batteries - you'd have to remove the passenger compartment from the lower frame. The car is not designed for that form of "refueling".

No you're definitely wrong about this. I don't think they're widely available but the S is designed for drop in battery replacement

They did a live demo:

https://www.youtube.com/watch?v=H5V0vL3nnHY

Here's a closeup video where you can see the battery pack being lowered and another being pushed into place:

https://www.youtube.com/watch?v=HlaQuKk9bFg

That swap site was designed to arbitrage a loophole of California tax credits for EV's.

http://www.latimes.com/business/autos/la-fi-hy-tesla-battery...

Basically, if a battery swap facility exists and has done it on at least one car, then all cars in the fleet capable of this battery swap can enjoy the extra tax credits, even though said facility is not open for actual use and has no plans to employ swapping on the other vehicles.

There's no reason to pull out the True Believer ad hominem. You have valid points, but it also wouldn't be the first time that a long list of drawbacks with new technology are resolved one by one until it becomes obvious to everyone that the new way of doing things is better. I think that battery-electric vehicles are in this category, and that it will be tough for competitors to catch up once the "this is obviously the better way" point is reached. This is obviously a matter of opinion, but it certainly isn't clear-cut.

I'd also claim that the drawbacks you mention don't merit the word severe, given that Tesla is able to sell ~100,000 vehicles per year with these drawbacks. They are certainly disadvantageous, but not to such a degree that they are a limitation to more than a few precent of the use cases. Also, there are advantages to electric vehicles - e.g. less routine maintenance, fewer moving parts to service, lower operating costs, less noise and emissions. So to a degree it evens out even with the present limitations.

100k/year compared to 17m cars and light trucks sold in the US in 2015 is a drop in the ocean.

This seems to be another example, particularly on this forum, of "living in the bubble". The bubble in this case is urban living in California, a situation with relatively low ranges of driving with temperate weather.

Again, I'm not saying it won't happen. I'm simply pointing out the obvious: it's got incredibly far to go and there are enormous challenges to overcome such that, at a minimum, it's going to take a long time.

Dominating their niche will be a useful outcome. If light vehicles in cities become predominantly electric, that is a useful outcome and I look forward to it for the reduction in air and noise pollution.

I think that "niche" covers about 70% of personal vehicles in the EU.

Edward Niedermeyer (the author of this particular opinion pieces) hates Telsa, loves Toyota. Pretty clear if you have followed any of his articles.

Same thing with reusable rockets. First it won't work, now it won't be economical.

These are moribund industries where nothing has changed since the 60s. After a while they've come to believe that nothing can change.

Lithium is more common in Earth's crust than lead and many other materials that are mined in significantly greater quantities.

the automotive industry has not changed since the 60s? Your just making completely factually inaccurate statements.

Point #3 has always been true. Tesla has been having difficulties scaling production or making ship dates for years, now.

I've yet to see anything that will convince me that they will ship 500,000 cars in 2018.

And SpaceX will never be able to launch and re-use a rocket. And residential solar will never be mainstream.

I get the point of the original article, production manufacturing at this scale is a whole different animal from making products at lesser volumes for early adopters.

But there's another way to look at it. It's another challenge. And so far, Tesla has been up to the challenge. And so has SpaceX.

It should be clarified that this is an op-ed. The author does not cite any hard facts, and the primary argument seems to be that the author believes that only Toyota's methods are the methods that work, and because Tesla isn't doing what Toyota did, but has ambitious goals, it will fail.

Yeah yeah. But does anyone here with a history of building something more complicated than a Rails app want to bet real money that Musk will put a Dragon on Mars AND ship 500,000 Teslas in 2018?

He's not exactly known for hitting his ship dates or volume targets.

What is exactly the penalty for missing ship date targets? Slight cost increases? Lower profit margins? Do they have difficulty with cash?

- Until you start shipping, you are burning money. Hit 0 and you die. Tesla is widely believed to need another round before it starts shipping Model 3, even if everything goes to plan. [1]

- US tax rebates for electric vehicles are phased out after a manufacturer has sold 200 000 of them. [2] The longer Tesla takes to start selling Model 3, the more it will have sold of the upmarket (hence less rebate-sensitive) S and X, the fewer price-sensitive Model 3s it will be able to sell at the rebated price.

- Competitors are not standing still. Who knows what else will be available in 2019 or 2020? And how long will those 400k preorders stand in the face of delays when there are alternatives?

[1] http://uk.businessinsider.com/tesla-to-raise-more-cash-2016-...

[2] http://www.greencarreports.com/news/1085549_when-do-electric...

People will buy an alternative electric car which is very likely to be available from other manufacturers by then.

I don't put too much trust into Elon time, but I do believe that if it is at all possible to achieve, Elon is going to will it into existence.

Are you saying that he'll get to Mars, but not on time? I'm positive he'll do it alright. Whether its 2018, 2019, or 2020, hardly diminishes the feat.

Agreed, but just a correction, if not 2018, then you don't get a launch window for another couple of years.

Reminds me of Louis CK rant "..cause now we live in, in an amazing, amazing world and it's wasted on the, on the crappiest generation of just spoiled idiots...". The dude disrupted 3 major industries and all we can do is complain that he is not hitting ship dates or volume targets let's just stone him for this horrible transgression.

Four! Look at what PayPal did for online banking... They pioneered it.

The issue is that people will jump ship especially since the reservation fee is refundable. By 2018-19 other manufacturers will have alternative electric vehicles ready to ship.

It's been years and there is no credible competitor to the Model S. Maybe it will be different with the Model 3, but until an equivalent car is on the road I'm not sure I agree with your statement. The Bolt - a nice car, and I've test driven it - is just so clearly sub-par compared to what Tesla is aiming for with the Model 3 that I would not consider it a true alternative.

So Tesla has two problems, according to the article:

- They can't ramp up production fast enough to keep up with the demand indicated by pre-orders.

- Pre-orders are likely to be overstating demand substantially, since they are refundable and therefore vulnerable to speculation.

...Either one of these problems by itself could be a threat, but both together seem somewhat less worrying.

If you interpret "can't ramp up" as "can't ramp up without extreme cost/risk," one could envision both being a problem. They could go all in and fail to see the demand they were expecting.

Tesla's two top manufacturing execs quit just before Musk announced that he wanted volume production of the Model 3 in 2018 instead of 2020. You can imagine the discussions that went on. Perhaps in time details will surface.

Tesla now faces transitioning from a low-volume high-margin manufacturer to a high-volume low-margin manufacturer. That's hard. HP, IBM, and Xerox tried that, and didn't do too well; all eventually exited the low end and downsized. Even Intel is exiting the low end; they're giving up on the tablet and phone market, as a YC article pointed out a few days ago.

Tesla is trying for good, fast, and cheap at the same time. Tesla's track record is that they can do "good", but late and at a high price. (Space-X has "good" and "cheap", but is about two years behind announced schedules. Their competition is even slower, so that's OK.)

Tesla looks like they're doing a good job of manufacturing, but their Fremont plant has 3,000 employees for 50,000 cars a year. That's 125 labor hours per car. They can't make a $35K price point with that. The big manufacturers are in the 15-30 hour range.[1] That's the big problem. Tesla needs to learn how to reach Toyota levels of production efficiency, and they have a long way to go.

Making 500,000 cars a year isn't in itself an obstacle. The Mercedes A-class production line cycles every 40 seconds. That's 720 cars per shift. If that line ran 2 8-hour shifts (you need a shift for maintenance) and 7 days a week, it could build over 500,000 cars in a year. That's just one production line.

It's getting that production line built, working, thoroughly debugged, and running at a low labor cost that takes some time. Tesla is going to need people who've done that before.

[1] http://www.autonews.com/assets/PDF/CA2018861.PDF

Ways Tesla can scale up:

1. Add a lot of automation. Costs would be excessive.

2. Add a lot of people. Cost would be quite high.

3. Improve production process. A 10x improvement in production rate in two years would be incredible to see.

4. Add another factory. See points 1 and 2; also could not happen in 2 years.

5. Partner with another OEM. This seems to me the most viable way to Musk's goals in his timeframe. We shall see...

6. Suspend building Model S and Model X vehicles; concentrate all resources on the Model III.

Good and insightful criticism. Thanks for the input :) A refreshing alternative to knee-jerk dismissal.

Something they don't mention is lithium supply. According to the Financial Times[1], that's a big problem for them:

> Tesla will need about 24,000 tonnes annually of lithium hydroxide, according to Benchmark Mineral Intelligence, out of a market last year of 50,000 tonnes.

> more than 70 per cent of which is found in Chile, Argentina and Bolivia. Due to growth in demand for electric batteries, the global lithium market is approaching a shortage, with no new supply coming on stream next year, say analysts. At the same time, battery factories being built in China are set to increase demand for lithium.

> So far two companies have announced supply deals with Tesla, and neither is expected to produce substantial amounts of lithium until after 2020.

> “The current strategy seems to be no direct investment but leveraging the Tesla name by signing ‘contingent’ contracts at unachievably low prices with junior mining companies who have never produced lithium chemicals,”

> Albemarle of Baton Rouge, FMC Lithium of Philadelphia and SQM of Chile, the big producers of lithium, have not been willing to supply at the prices Tesla has demanded.

[1] http://www.ft.com/intl/cms/s/0/4a924a64-99df-11e5-987b-d6cde...

(To get around paywall, google "lithium supply tesla" and take the top result.)

The 70% appears to be proven reserves (at least, I don't see reason to interpret it otherwise).

The more interesting question is where supply can easily be developed. The much larger reserves elsewhere don't matter if a different mine is able to easily increase production to meet the demand.

Lithium is abundant, the question is just how well it can be extracted. For example, there is more Lithium in sea water than humanity can ever use to build batteries, but we still need to commercialize the extraction. So it's not that there isn't enough of the stuff, it's that there currently isn't that much of a need for it, so only the most easily accessible reserves are being tapped. That might cause supply issues in the short term, but long term, there's no shortage.

In the long term future of humanity, sure. I'm talking about the next few years for Tesla, since that's the topic of OP's article.

Production is increasing, but slowly, and even if Tesla were offering a price sufficient to motivate faster increases, it takes time to bring new projects online. Aside from building facilities, you eat a couple years just waiting for brine to evaporate: http://www.bloomberg.com/news/videos/b/3abb52e4-ea59-4ed9-85...

  battery Li content =  80g / (kW hour) [1]

  Gigafactory total projected capacity 2020 = 85 GW hr/year [2]
  Gigafactory lithium metal demand = (80 g / (kW hour)) (85 GW hr/year) = 6800 t/year
  Gigafactory lithium hydroxide demand (anhydrous) = (6800 t/year) (23.95 g/mol / (6.941 g/mol)) = 23500 t/year
  Gigafactory lithium hydroxide demand (monohydrate) = (6800 t/year) (41.96 g/mol / (6.941 g/mol)) = 41000 t/year
  (FT article was evidently talking about anhydrous lithium hydroxide)

  price lithium hydroxide = USD 14000/t [3]
  price lithium carbonate = USD 10000/t [3]

  price lithium (from hydroxide) = USD 14000/t (6.941 g/mol / (23.95 g/mol)) = USD 4060/t
  price lithium (from carbonate) = USD 10000/t (2 6.941 g/mol / (73.89 g/mol)) = USD 1880/t
  (N.B. just cost of raw materials, not including cost of lithium extraction from salt by electrolysis)

  Gigafactory demanded lithium cost (from hydroxide) = (USD 4060/t) (6800 t/year) = USD 27 million/year
  Gigafactory demanded lithium cost (from carbonate) = (USD 1880/t) (6800 t/year) = USD 13 million/year

  Estimated world lithium reserves = 14 million t [4]
  Gigafactory proportional use of lithium reserves = (6800 t/year) / (14 million t) = 0.05%/year

  World production of lithium (2015) = 32500 t/year [4]
  Gigafactory proportional use of lithium production = (6800 t/year) / (32500 t/year) = 21%

The FT article's estimate of 24000 t/year checks out for anhydrous lithium hydroxide. I do not know if 50000 t/year is an accurate number for worldwide lithium hydroxide production in 2015, but lithium hydroxide is not the only source for lithium and this would imply a worldwide supply of lithium of just 14500 t/year, which is contradicted by the figure of 32500 t/year from the USGS [4].

The estimated lithium demand of the Gigafactory (which matches the FT article's figures) are for the Gigafactory's total demand (both Powerwalls and cars) with the factory at full operating capacity in 2020. There is a number of years yet for lithium production to scale up.

Even taking the Gigafactory's projected demands in 2020 against lithium production in 2015, it requires 'only' 21% of the market. Tesla would be a very large customer, perhaps the single largest in the wourld, but Tesla wouldn't exactly be required to corner the entire market.

Supply and demand dynamics dictate that with increased demand, price will increase. The cost of the raw lithium salts would be on the order of USD 20 million/year for Tesla. Even with a price vastly inflated from here, it would be a small cost against Tesla's operating income.

A single Gigafactory operating at full capacity would take over 2000 years to exhaust the USGS's estimate of world lithium reserves of 14 million t.

[1]: http://electronics.stackexchange.com/a/34504

[2]: http://www.zdnet.com/article/panasonic-to-pour-billions-of-y...

[3]: http://www.abc.net.au/news/2016-01-13/lithium-graphite-the-n...

[4]: http://minerals.usgs.gov/minerals/pubs/commodity/lithium/mcs...

You inverted the mass ratio of Li-mass to compound-mass in the calculation.

price Li per ton from LiOH:

    23.95 [mass of LiOH] / 6.941 [mass of Li] * $14000 [$/ton of LiOH]
    $48,310/ton [of Li]

price of Li per ton from Li2CO3

    73.89 [mass of Li2CO3] / (2*6.941) [mass of Li] * $10000 [$/ton of Li2CO3]
    $53,230/ton [of Li]

Damn, you're right. Thanks for pointing that out.

Unfortunately, I don't seem to be able to edit my earlier post.

That changes some of the numbers substantially. Cost of lithium salts at 2015 prices to sustain full Gigafactory production becomes about $USD 340 million/year, rather than the $USD 20 million/year I initially calculated.

The figures about volume of supply, demand and reserves, stand.

Sometimes I think it would be worth putting in "dummy parameter" units to stop mistakes like this from happening with dimensionless ratios.

Anyway, thanks again for pointing that out.

Based on title, expected to see the article analyze lithium mining capacity, packaging constraints, limits to manufacturing efficiencies, P&L analysis of growth capital requirements, etc.

Got "Toyota is better" instead.

"shows how badly Musk fails to understand what he's up against"

Lines like this is why writers write and doers do.

Tesla has yet to transition from a product first to a manufacturing first kind of company. Entry level cars are much more shaped by manufacturing possibilities rather than consumer demand. It's not unreasonable to assume that the Model 3 will be sold at a loss and then it will be a question about how much longer investors are willing to pour money into Tesla rather than how many cars they can sell. I'd buy a Tesla - it's a bargain. Unlike the stock.

Worst case scenario, wouldn't it be conceivable that the guy who built a car and space rocket company from scratch now creates/buy a lithium mining and/or battery making company to be able to supply Tesla different projects (cars, powerwall)?

And make it as autonomous as possible.

Tesla is probably pursuing a partnership with a company like Toyota that can already mass-produce autos. If GM can do it, why not Tesla? (The author didn't mention that possibility, which makes me suspicious that he has an axe to grind against Tesla.)

What Tesla has, are Musks shoulders. He went pretty deep and pretty low to make this happen when it was nothing. Not to make him a supernatural being, but now that his ventures are stable and growing I expect him to be able to handle pressure and find solutions way easier than when Tesla almost went bankrupt.

This is probably the only direction available to Tesla. They can grow or lose market share over 5 years as the legacy automakers catch up.

The biggest threat to Tesla now is that their logistics teams will be able to source sufficient materials. I think getting the manufacturing lines up and running will be a challenge, but one Tesla has met before.

Financially can Tesla do it? I'd say yes, and they probably won't even have to do any fundraising to do so. They will probably be able to fund it completely through debt instruments. They are a good bet by any measure at this point.

Most silicon valley startups try to learn from the Toyota manufacturing process and improve on it. While Toyota got stuck, Tesla gets rid of the most inefficient part of manufacturing process: humans

Humans play an important role in the continuous improvement process.

Absolutely. Continuing with the Toyota example, more direct human involvement with the manufacturing process has been reintroduced [1]

[1] http://www.bloomberg.com/news/articles/2014-04-06/humans-rep...

It's probably the way not to do things if you want to get to affordable supercars. The only way is to program robots to manufacture the cars from start to finish, and the development should be in software (teaching the robots to work better)

Headline is misleading. Current source headline: "Tesla Needs More Than Elon Musk". Vs HN headline: "Tesla will not be able to scale its manufacturing capacity"

Disclaimer: I work for a Tesla competitor, any opinions are solely my own.

I've noticed several questions on this thread about manufacturing. Since I have worked in auto manufacturing, I will share some thoughts in the following areas: EV Assembly is Simpler; Humans are inefficient; Musks' Desk by the End of the Assembly Line; and Tesla is Smarter than Toyota.

This is intended as a VERY high level overview, expect many details to be glossed over.

EV Assembly is Simpler: Electric vehicles are slightly simpler than Internal Combustion engines. This does make EV Assembly slightly simpler... but not much.

Automotive Assembly plants consume parts and make finished vehicles. Assembly plants have three main areas: Body, Paint, and General Assembly. Tesla's Fremont plant has these 3 areas.

What is because Tesla makes EVs? Not the Body Shop, not the Paint shop. Is General Assembly simpler? Not by much. GA is split in two areas - Trim and Chassis. Trim no less complex on an EV than ICE. Part of the Chassis process is simpler.

So, part of 1/2 of 1/3 of the Assembly process is easier in the Assembly plant; call it 10% less complex (1/6 = 17%, but only part of that 17% is simpler).

Humans are Inefficient (also, Four Function Robot):* Efficiency, and engineering is all about trade-offs. In my opinion Musk will not be able to automate his way to 500k vehicles per year.

Humans are often perceived as inefficient because: software engineers (and most of the people reading this) tend to value their time highly and strongly dislike repetitive tasks. The perception is that if the task is automated, it is a better use of a persons' time. This can be true. This is why all auto OEMs have some automation in manufacturing.

Humans, however are better in a number of ways: humans are great at learning new tasks quickly and cheaply, humans require much lower capital expenditure to hire (they may be more expensive in the long run, but 2 years is not the long run)

Automation can also be inefficient depending on the parameter chosen. At one point Tesla had a multi-function robot that changed it's "hand" for each vehicle to perform different task. Changing the hand takes time, which Tesla cannot afford if they want to build 500k vehicles/year. Yes, you can simply add more robots, but that costs money.

As Tesla has been setting up its manufacturing process, it may have in mind future expansion and speed-up; but it has been solving current problems. Automation in particular can be troublesome to speed up (most definitely following the 80/20 rule). Portions of the assembly line may have to be re-structured to allow parallel operations, which can be expensive and time consuming.

Also, the assembly operation(s) in question have to be engineered with automation in mind (else it will be amazingly expensive). This is entirely possible, but I don't see a major breakthrough in 2 years alongside a major vehicle launch.

Musks' Desk by the End of the Assembly Line (and Tesla is Smarter than Toyota): This item in particular makes me shake my head in disbelief. On a recent earnings call, Musk said that his desk is next to the end of the production line [desk]. This suggests that Musk is attempting to inspect his way to quality. Do you know who tried to inspect their way to quality? GM (among many others). Do you know who has a good reputation for quality? Toyota. [quality]

In addition (and more pertinent to the speed-up question), quality has a quantity all its own. If you use high quality parts, you can put them together more quickly and more reliably, and thus in higher numbers. If you produce high quality vehicles, you can ship them more quickly, and again, in higher numbers.

In other words, the defects are not being added to the vehicle at the end of the assembly line, and they cannot easily be removed at the end of the line. Can you test your way to high quality software? Can you inspect your way to high quality vehicles?

[desk]: http://qz.com/676127/no-need-to-worry-elon-musks-desk-is-rig...

[quality]: 1. Musk is certainly doing more than just inspection, but his words send a very naive message. 2. GM and other companies perceived as low quality have actually caught up to, and surpassed Toyota. http://www.jdpower.com/press-releases/2015-us-initial-qualit...

It's interesting because the software development strategy of break things fast is based off of agile methodologies developed by Toyota. If any place is primed to be the birthplace of a car company that emulates Toyota, then SV is it. And there's no reason tesla can't be one of those companies.

whoa whoa whoa... the foundation goal to which Toyota's strategies have been focused for the last 50 years have been explicitly towards improving quality of the vehicle once it leaves the factory... The time to market focus at the cost of quality that is a rational viable model in software is the antithesis of the toyota way...

how does tesla compare to byd?

http://www.europeanfinancialreview.com/?p=5603 http://www.byd.com/news/news-313.html

i think byd had pretty big claims when they started out into cars too..but went quiet after awhile...but somehow they'd survived and its been more than 10yrs since they started...

What a boringly one-sided article.

This article is wrong headed in one very important way. Electric cars are fundamentally a much BETTER way to make cars. And that's why Musk (and other electrics with decent engineering) can count on selling all of his production.

Reply on news.ycombinator.com