- Incumbent auto giants entrance diminishing Tesla's margins.
- Incumbent expertise and brand that diminishes Tesla's appeal.
- First mover disadvantage: rest of auto industry sees large initial ROI by piggybacking off Tesla
GM can't be the only guys racing to bring Tesla experience to it's existing brand followers. Luxury makers like BMW, Benz will woo the lion share of the market Tesla originally pioneered. First movers rarely do well here unless they've cemented their brand position in the luxury space such as Ferrari, Lamborghini etc.
tl;dr: When the status quo catches up to Tesla, it will put downward pressure. Tesla's winning strategy should be to limit quantity and artificially inflate unit price. It is not equipped to take on dozen automotive brands in the long run.
You mean like how all the cell phone makers caught up with Apple's iPhone?
Not saying TSLA has no risk: for one thing, they still have not figured out how to scale up production. But in general, I think we tend to overestimate the incumbents' advantage. Most incumbents have too many old interests that hold them back. There is an article [1] about German automakers' unwillingness to commit to EVs because the profit from ICEs are way way higher than that from electric ones because of ICE's complexity.
(Similarly, this is the reason why most of the wind farms in China are not connected to the national grid, because the existing power plants do not want to shut down and lay off workers.)
Agreed. What strikes me as perhaps noteworthy is that I can remember when Hyundai started selling cars in the United States, and it took quite awhile (from memory, until the early 2000s) before anyone was taking them seriously. Meanwhile, Tesla seems to have become entrenched in the mindshare of the American populace much, much more quickly (in the same way that the iPhone did relative to Android).
Tesla may or may not be around to reap the rewards of their popularization of the electric car, but they've definitely made an impact, and done their part to push automotive thinking forward by at least a decade IMO.
Engineering a reliable car is hard. Even Nissan and Toyota, who spent decades wiping the floor with cars from the big 3, struggled when it came time to apply that expertise to building full size trucks.
Yes, there is some institutional inertia that holds back the incumbents. However, there is also a huge store of knowledge about what to do and what not to do.
Here's a somewhat relevant story from my time at Boeing. When the 737-7/8/900 was being designed, some young bloods looked at the vertical stabilizer and said "Oh my God! Look at all this useless structure! We'll use cutting edge FEA and CATIA and show those old pencil-and-paper slide-rule jockeys how it's done." Everything looked great, and they started building the new aircraft in Renton. Except that during flight testing, they found that the new vertical stabilizer assembly was cracking. They had to stop the line (very expensive), remove assemblies from built aircraft (very expensive), and pay penalties to airline customers for delivery delays (astronomically expensive). If memory serves, they had to take a $1B writedown. There were also public firings of high-level executives.
Name a major smartphone maker pre-iPhone that's still a major smartphone maker now. Being first with an inferior product (Tesla certainly isn't in that category) is irrelevant to the discussion at hand and the parent poster didn't claim Apple was first, just that they lead the market (overcoming all the incumbents at the time) in the early days before Android caught up and passed them.
Tesla is inferior to all luxury brands in all aspects except being electric and self driving. When compared to Volvo, Mercedes and their ilk, Tesla are behind on prestige, interior styling, after-sales support, economies of scale, supply chain management, navigating world wide bureaucracies, etcetera. It remains to be seen if Tesla will catch up on these faster than the other brands can catch up on self driving and electric vehicles.
Right, except in the hard tech areas. I suspect it will becomes a bigger differentiation when setting the standard. Focusing on chrome would be a waste of immediate effort. It makes more sense to work on the commonalities of the energy infrastructure that augment the existing landscape than on what chrome goes best with the seat's heated-leather-ass-warmer. The luxury brands won't have the energy generation and storage solution to match. The competition has an enormous supply chain problem that's just getting worse unless they jump on board with Tesla's charging standard.
In Europe they have standardized on a different charger design, I don't see how charging can be a major differentiator once electric cars become common. A charging station is extremely inexpensive compared to a gas station and those are ubiquitous
Hi, can you give me the sources/reasoning behind your statement?
I have first hand sales experience with luxury brand vehicles (read: RR/Bentley price class). Technology oriented US customers love Tesla. Sure, top notch luxury is nice, shiny and gives a unique experience, but at the end it is a different kind of experience from that a Tesla customer seeks. At the end, they buy both.
I've driven many cars from most major luxury brands (including Porsche, Mercedes, BMW, Ferrari, Bentley, Jaguar, Lamborghini, even Bugatti). At it's price point, there's no other car I'd personally rather drive. Maybe my characteristics for a car are different than yours. But as far as comfort, performance, feature set, and styling I vastly prefer the P100D over any other car in its price category (aka $100-150k range).
0-60 time is a little important to me (but no, I don't need Tesla's level stats). Getting on a freeway and merging in traffic is an example of where it's useful. I'm not one of those drivers who thinks it's ok to not be going traffic flow speed by the end of the entrance ramp, so having an electric car that can easily do that is something I value. Many hybrid / EV vehicles strain to do it at times in my experience (I've test driven most of them on the market today).
Apple was the first company to make the smart phone as we now know it. The entire concept across the smart phone spectrum today, is directly derived from what Apple did with the first few iPhone iterations - in every possible regard. Every competitor that came after for years did little more than copy the iPhone's features.
Tesla is a vertically integrated company, so they have squeezed all the margins out of their supply chain. The big automakers all share their supply chain and have to feed it with margins at every level. Same story with SpaceX - vertically integrated.
The whole integrated/shared thing has been going back and forth in different industries at different points in time. Apple has been getting more and more integrated recently whereas earlier wisdom in PC industry was to use shared components.
Musk is betting that it's time to integrate the automotive industriy. If he is right, Tesla will balooon to the size of Apple. The electric part is not the most important here.
Personally I think Musk doesn't think this way. I think he sees the secondary aspects of Tesla's work on cars (e.g., the battery plants and larger infrastructure of the electric car industry and positioning on standards bodies) as the actual big fish to hunt for.
Tesla cars are nice, but they didn't need to make the entire pipeline. They're doing that because they want to reuse the pipeline and tech in other industries.
I believe Musk means it when he says he wants to wean humanity off of fossil fuels. The big fish you talk about are definitely means to this end and not the goal itself.
> Tesla is a vertically integrated company, so they have squeezed all the margins out of their supply chain. The big automakers all share their supply chain and have to feed it with margins at every level. Same story with SpaceX - vertically integrated.
Vertical integration doesn't guarantee better prices just because you don't rely on third party suppliers. There are other benefits, such as monopolizing a certain part of the supply chain (like batteries in the case of Tesla) but it also requires very high investments (Gigafactory). Those investments can pay off in the future but can also turn out to be very expensive mistakes - you basically wager long term pay offs against flexibility.
Tesla uses the same Tier 1 suppliers as any other automaker, so I don't really see why the Tesla production would be more vertically integrated. They have their own batteries but so do traditional automakers have their own ICEs and other key technologies.
Everyone is missing the point on vertical integration. It's to iterate faster. Even in atoms not bits? Yes!! This is nearly the gigafactory motto, the machine that builds the machine etc.
>GM can't be the only guys racing to bring Tesla experience to it's existing brand followers.
This has been pointed out many times on these boards, and I agree with it.
Tesla (Musk) got this EV ball rolling, for real, and deserve immense credit for that. But it sounds too often here that Tesla has "won". Building cars is hard, capital intensive, and there are many massive companies diving into this. And despite what you may have heard, companies like Ford, GM, Fiat aren't staffed with idiots.
Absolutely. Big autos have huge engineering know-hows as well as resources to poach talent.
One thing that may work for in favor of Tesla though, is the fact that self-driving car will fundamentally shift people's relationship with cars. Out goes the needs to have best-handling, best crash protection, best-feedback providing cars.
Value of those engineering know-hows will be greatly diminished because computers will be doing the driving. Instead, there will be more emphasis on comfort, software, and efficiency.
Electric drivetrains alone were never really regarded as a major competitive threat to the big carmakers. What has lit a fire under their ass is is the combination of electric, autonomous, and on-demand. This is disruptive, this scares them.
Google's autonomous driving efforts alone weren't taken too seriously either, but when rumours of Apple's interest in the space began to swirl, that's when the carmakers (including Tesla) got the message and suddenly all started scrambling to come up with a strategy to develop an autonomous OS. Fast forward 3 years and here we are, with stodgy old GM in a position of confidence.
Tesla's own charger network also presents a slight advantage for potential buyers, other brands [for the moment] rely on home chargers, dealership sockets and third-party vendors like Chargepoint or Volta.
Tesla has also opened the patents for their socket and charging designs. I can see a future where Tesla isn't necessarily selling a ton of cars, but is licensing technology and selling infrastructure (such as access to charging stations). This makes the purchase of Solar City make even more sense- they may care more about moving into the utility market than the car one.
I strongly suspect that's the strategy. Reusing supply chain/parts to build competitive new energy infrastructure, licensing commercialized technology to major manufacturers in order to fund R&D, and using early adopters/status products to "beta test" the next generation.
The battery cells inside the Gigafactory are created by Panasonic. So far it looks very unlikely that the Gigafactory will be able to produce battery cells at a substantially reduced cost when they're relying on subcontractors for all parts. Innovation in batteries is really hard. We've had lithium-ion batteries for nearly 50 years now, and they're just not improving that quickly. We'll get 5% a year improvements for the next decade if we're lucky. The Gigafactory won't change this. So I don't see how the Gigafactory will be a lasting competitive advantage.
No source, but I saw some comments here or on Reddit saying that getting batteries in the large quantities you need for EVs is very difficult. Tesla's Gigafactory allows them to not have to compete for that supply, which may give them a price advantage over others buying from Panasonic.
Battery manufacturers are already preparing for the electrical car revolution. Even if li-ion cells are supply constrained today, they won't be 5 years from now.
The gigafactory and charger network give Tesla some barrier to entry but I don't think the costs will make a giants blink.
Even more scary is collusion of giants trying to kill Tesla. Say the Japanese automakers group together to share the risk and costs of building the same infrastructure.
My primary concern is the margin per unit. Only a handful of automakers today have been able to sustain their margins, namely Porsche which supposedly makes the most margin per car. If you wonder outside the luxury segment, you find razor thin margins and heavy competition.
Tesla and Elon Musk deserves a big award for bringing the EV market to people who succeeded in capitalism. But as you saw with Ford, just because you were first won't guarantee your place. Heck, without bailouts, there would be very few American automakers left.
And this is really where my argument for maintaining high margin per unit is essential for long term survival. That can only be accomplished by moving upwards, not downwards. The accumulative cost & risk of selling 2.5 model 3 vs 1 p90D can't be ignored.
Having said that the market seems undecided between the bears and bulls on the upcoming Tesla earnings report.
> But as you saw with Ford, just because you were first won't guarantee your place. Heck, without bailouts, there would be very few American automakers left.
While ford had the cash to survive the Great Recession without a bailout, it would have not survived the collapse of the supplier networks when GM and Chrysler went under.
Would these purchasers have completed the purchase in time to keep the suppliers in business? If GM folds then the part supplier may feel fairly confident that someone else will purchase the assets and start ordering again, but their bank who lends them the money they need to weather the storm may be far less confident that the deal gets sorted out fast enough for the supplier to stay in business. Now multiply that by a hundred.
No way the supplier network would have survived a dual bankruptcy.
Perhaps one of these [1] in the boot would reduce stress. It gives about 4 miles of range per hour of charge [2], so only for reducing stress, not for actual planned use!
I owned a Leaf for 2 years. "Winging" it really, really, sucks. Especially when you have your family with you, who don't share your enthusiasm for electric vehicles. :D
I just bought a Volt for this reason. My wife will be able to stay electric 90+ % of the time, but this weekend's 150 mile round-trip out of town and back for her will be stress free. I'm not sure even being in a Tesla would be as convenient since I don't think there's a supercharger directly on the route.
Yeah, that's a good point. For some reason I was thinking of the Bolt with more like 200 mile range, which could still make this trip, but it would be close once you add in the bit of driving she did at the destination.
But I think the larger point is still valid: the charging infrastructure is still not nearly as widespread as gas stations and charging still takes longer than re-fueling. For us, the compromise of a range extender is still the best of both worlds. We can run electric most of the time, but we can still use the Volt as a "regular car" for road trips.
GM has a long track record of fumbling new products. I expect Tesla will suffer from competition, but it will be BMW or Toyota that does the damage.
The big advantage Tesla has, as far as I can see, is the charging network. Someone else is going to have to bite the bullet and put up a few thousand charging stations.
There are plenty of non Tesla chargers here in Kansas City. I can charge my Nissan Leaf on a chademo charger in 1/2 hour. So I'm wondering if Teslas own charger network makes that much difference.
First, they're not organized into a coherent network. When making a long trip, you may need to go well out of your way to hop from one CHAdeMO location to the next, or just not be able to do it altogether. Tesla's network is not entirely ubiquitous yet, but it covers 95% of cases now.
Second, somewhat related to the first, they tend to be installed in small numbers, often just as single units. That makes it almost impossible to rely on them. If you arrive at a single CHAdeMO unit and it's broken or in use, you're screwed. Tesla Superchargers are usually installed of groups of 6 or more. For site-wide outages, the car will tell you and route around the broken sites.
Third, they are ssllooww. Charging in half an hour sounds nice, but that's a Leaf, with a tiny battery. Plug a Tesla into a station like that and you're looking at a couple of hours. Most CHAdeMO stations max out at 50kW or less (I believe many of Nissan's are only 20kW!), while Superchargers do up to 120kW. Recharging time on road trips is generally acceptable with Superchargers, but with CHAdeMO you're looking at doubling or tripling that time, and it'll start to become a significant portion of your total travel time.
Most people want to be able to make long trips. It's the #1 question I get about my Tesla. So far, no other manufacturer has shown anything capable of long trips. The Bolt could do them, if it had a reasonable charging network and could charge quickly, but it doesn't and it can't.
Chademo is about 1/3rd the speed of a supercharger in practice. I've never seen a full power Chademo and most have been 40-45kW. I get 120kW each time I hit the supercharger near me at <35% battery.
As a Model S owner, I may be biased, but Tesla's competitive advantage goes beyond electric drivetrain and self driving. Other things that work for Tesla that owners love and makes it very difficult to switch
- Supercharging infrastructure. My road tripping has actually gone up because of this
- A software engineering mindset, which results in regular infotainment updates, constantly improving features and a vehicle that gets better with time
- No dealer network - which leads to a customer first mindset.
Incumbent automakers are cash cows, not startups. Tesla will spend all of it's money on the Gigafactory. GM will destroy new product lines that threaten it's current profits.
Remember, incumbent auto makers are not fast moving or innovative at all. They are all cash cows. Although they can predict that a move to EVs will keep them alive long-term, such a move will always be done with a small fraction of GM's resources and labour in the coming decades, and Tesla with Elon's burning desire to destroy them are probably going to win.
Remember, GM can't sell these two products on the same lot: One which is a cash cow, and one which only exists to combat some distant threat of GM becoming obsolete.
Also you're forgetting that Tesla cars are just better in every sense. Better brand, faster, longer range, novelty...
There's a lot of companies out there, and dismissing them outright is the same brand of arrogance that lets startups win over incumbents, but reversed.
> better brand
GM/F/Honda/Toyota/etc are reliable, BMW/Audi/Lexus/etc are luxurious and have a package that's been very carefully tuned
> faster
Doesn't matter to 90%
> longer range
Debatable, also price constrained
> novelty
Not what you look for in a massive purchase that you want to last 10+ years
Just so we are clear, the Bolt is a hatchback, not a tiny hatchback. At 56 cu ft, it's a hair 2 cu ft) less than a Model S. The footprint isn't significantly smaller than that of the Model 3.
Bolt's website calls it a subcompact car. Despite what it's actual volume is, it looks like a small car and is classified as one. The Model S is classified as and looks like a luxury full size sedan. Yes, the Model 3 is comparable to the Bolt, but it's not being shipped yet. So comparing the distance of a Model S and a Bolt isn't a fare comparison as the Model S is bigger and heavier, hence my earlier comment in reply to the comparison. If you think comparing the EV range of a 3,500lb subcompact Bolt as being good compared to a nearly 5,000lb full sized sedan then fine, we can agree to disagree.
Musk has been working on increasing battery production for years. I don't see other car manufacturers outpacing Tesla for the simple reason of not having enough batteries.
I think Elon Musk is ok with that. Actually, I believe that is why he started Telsa, and opened up a lot of their patents at some point. He wanted to kick-start the electric market, not have a monopoly in it.
The talent at GM, Ford, Mercedes, BMW, Porsche, Toyota, Nissan, VW, etc. is immense. The only factor for those companies is focusing it, which all of them have proven they can do.
What Tesla has, is talent + a big headstart. The talent part is not unique at all. The clock is ticking rapidly on the headstart, and Tesla is well aware of that.
I think the counter-argument would go something like "Tesla is a technology company like Apple, leading to superior product design and user experience due to more talented people and a high level of vertical integration. Thus, in the actual long-term competition of Uber/Daimler vs Lyft/GM vs Google/Ford vs Baidu/BMW (already collapsed) vs Tesla, Tesla will win due to its ability to best develop a digital vehicle experience at an affordable cost, fastest.
As an example of software advantage, Tesla is currently getting more real-world training data for autonomous vehicles than every other company combined.
As an example of hardware advantage, Tesla will be producing more lithium ion battery capacity from its gigafactory than the entire world could produce a few years ago.
As an example of product advantage. Well, have you been inside a Tesla?"
Or something like that I think... :P
But who knows, maybe Faraday Future will reveal a pixie-dust powered anti-gravity sphere that will make everyone else irrelevant.
I don't think Tesla has any sort of talent edge in the self-driving car space. Their original autopilot was mostly built on tech from MobileEye and from what I've heard Tesla doesn't pay great either, and unlike SpaceX their mission is not unique.
Having a network of cars collecting data is great, but they may not actually be able to collect all of it simply because it is super high bandwidth, whereas those with test cars can stick a rack of disks in the car and collect all of the data.
Does Tesla not pay well by industry standards, or is that when comparing them to software companies? Seems like they work their employees very hard, though, so maybe they lose out on talent– on the other hand there was a time when Apple worked its employees hard and they were still able to retain top people.
Regarding the data collection, that's a great point. I hope that's not the case though, as I would think the data is far more valuable than the bandwidth it's carried on. Though, I suppose the capacity simple doesn't exist to send it over the air the way you can just store locally and physically offload. Then I guess it comes down to whether the compressed data from an enormous number of vehicles still gives a significant advantage.
I own a Bolt, and I like it a lot. I have no desire to spend 4x that on a Tesla that is simply overspec'd for a simple commute car. I derive my self-esteem in other ways than fancy cars.
Maybe I should have waited for the self-driving version.
Why are people so confident that full-fledged self-driving will decrease ownership? I get that cars are usually parked, but IMO self-driving could turn driving into a leisurely experience which would increase my desire to do it in a personalized/homey space, increase my desire to have a long commute, and increase my desire to drive purely for leisure as well (driving around country roads reading).
Because the price differential will make ownership a luxury.
In fact, your comment alludes to that. For you, a self-driving car is an opportunity to make a leisurely, personalized, homey space. That's wonderful, if one has the money for it - and many people do.
However, let's take a step back and look at the millions of people who say, "I want a car that gets me from A to B". They're not looking to own a leisurely, personalized, homey space. They're looking to get places, hopefully at the cheapest price and in the most reliable way. They're not looking to own something.
You note that cars are often parked. Given the low utilization of vehicles, it stands to reason that renting time in a self-driving taxi should be significantly cheaper than owning one. If that self-driving taxi can serve the needs of 3 people instead of one, you only need to pay for a third of it. Even if the profit margins are an outrageous 50%, you'd be paying 150% of 33.3% or 50% of the cost of owning a self-driving car. However, the pricing is likely to be even more favorable. If owning a car costs you $500/mo and you can get your transit via self-driving taxi for $150/mo, there's a huge cost incentive to switch.
I think that's the big thing. The cost advantage is going to be quite stark. Owning a car will cost many times what getting driven around by a self-driving taxi will cost. While many people can afford a car today, it's certainly a significant budget item for many. Even for those that can comfortably afford it, it's safe to assume that many will ditch the cost for something cheaper.
Ultimately, our tastes as a society change as our society changes. As the cost of owning a car becomes sky-high compared to a mostly-substitute, many people will switch to it. Not everyone will. People spend up for luxury items all the time. But luxury items are, well, not commodities. Even if you can afford a luxury item, if you don't value that kind of item, you won't pay up for it. Plenty of rich people see a car as just a way to get from A to B. They'll take the cheaper self-driving taxi. Plenty of low and middle income people will be able to stretch their budgets better lowering transportation costs substantially.
Even if you're skeptical of a dramatic change, let's say I propose that 10% of people that currently own cars would ditch them for self-driving taxis. I'm not even suggesting that it's changing society, but a 10% dip in demand is quite significant to an industry.
I think you're right that there would be something nice about having a leisurely, homey ride. It could be a wonderful experience. But luxury automakers already try to give their owners a much nicer experience than someone gets in a Civic and there are definitely more Civic's out there. It's that, while a Mercedes S-Class is lovely, the Civic allows them to spend their money on other things (or afford transportation at all). Similarly, self-driving taxis will allow people to spend their money on other things or more easily afford transportation. You certainly paint a desirable picture and there are definitely people that will choose your vision. However, I think more people will choose the economy of self-driving taxis.
Cars made horses a luxury item. Horses were no longer common transportation. The car was cheaper and better. It certainly wasn't perfect in all ways, but enough that horses became a luxury item.
The key thing that privately-owned cars have over on-demand cars is reliability, and the illusion of control. I've owned an unreliable car (It had trouble starting, but would run fine once started). It sucked. Not knowing if the car would start each time changed my relationship with going places and doing the things I wanted to do. I had to budget extra time everywhere I went, in case I needed to jump the car or wait for AAA. On-demand cars will have many of the same problems, just due to demand surges, and traffic. Sure, you can get stuck in traffic any time. That's true, but you feel it less when you have the illusion of control of driving. When the traffic delays your inbound car by 5 or 10 minutes and you're just standing on the curb waiting for it, that feels more aggravating than when you're behind the wheel.
I also think there will be a significant fraction of the population that uses their car as personal, daily storage. My wife keeps a bunch of stuff in her car, and would be significantly impacted if she had to take it all in and out every time she went everywhere. Same thing with kids car seats, etc.
I think on-demand cars will probably replace second commuter cars for families, but there will still be many, many owned-cars as well.
That's a really good post and I agree that if the only added-demand is for people like me (btw, I am someone who only wants something to get from A to B with present cars) then that cannot possibly outweigh the likely large cost differential for the many many people who find cars difficult to afford.
I'm sorry for communicating poorly, but my point was more that it is not likely possible to predict all the consequences of a completely new technology and how behavior and economy will change in response. And I'm not even predicting anything myself about demand (beyond my own desires, which is dicey enough to predict) -- I am just curious where the confidence comes from when people state that ownership will be significantly reduced as an obvious fact when IMO that is certainly plausible but I don't have any reason for such confidence.
I don't think self-driving in and of itself will decrease ownership, but the combination of self-driving, flexible car sharing services, and ubiquitous ride sharing services will continue to make car ownership undesirable.
A car in a driveway or parking spot has horrible utilization rates. Ownership of things like cars has never actually made sense, given that they depreciate in value as soon as you drive it off the lot.
Meanwhile, more and more OEMs are moving into the car sharing space, and I think that will scratch the itch of wanting to drive a car for fun, or utility (need to pick up something large or be some where at a specific time).
One part of the answer is the prediction that ownership will move from personal ownership to fleet ownership.
There is also a factor that those predictions do not seem to be factoring in: That the need for a car is concentrated at commuter hours in the morning and afternoon. So the sales reductions will be tempered by the fact that many cars will be needed simultaneously.
Even if I didn't have some special access due to some relatives, I'd still have a Bolt this month as opposed to mid-next year?
By that time, Chevy will be shipping self-driving ones and according to the article they're going to NOT be doing that as a 40k addon.
I hope Musk has a very strong transition into a strong luxury brand. The simple truth is that there is a massive rush on personal transportation and all the small players are eager to sell to the Toyotas and Chevys of the world. These people are by and large expats from the early self-driving car projects.
'Thousands' is very committed. I wonder if they are doing this to make a statement, or if they feel rapidly scaling up to a thousands strong prototype test fleet is the most effective way to accelerate progress towards market ready L4 robotaxis.
As the analogy goes, just because it takes 1 woman 9 months to make a baby does not mean 9 woman working together can make a baby in 1 month.
Either way, it's very committed. GM isn't fucking around, they mean business!
I think it's in the same vein as Telsa's strategy of autopilot roll-out compared to Google's self-driving pilot program. 1000 self-driving vehicles running for 1 month can collect data at a significantly higher rate than 1 self-driving vehicle running for 1000 months.
It's the quality of the data and what you do with it that really matters. I think the benefits of crowdsourcing driving data to develop autopilot have been overstated. Waymo has pointed out that they are making the lion's share of their progress testing in simulators, where they can play out every conceivable variation of some strange or difficult edge case in a way that is not feasible in the real world.
One thing you can do with 1000s of fleet vehicles is begin untangling the logistics of running a publicly accessible robotaxi network, which will brings with it a whole new set of problems above and beyond the autonomous os itself.
A lot of people here are balking on the price for consumers. Looking it at another way is that the Bolt will be sold to ride sharing companies at 100k each and they would buy it in a heartbeat.
That question comes up a lot with big companies. The usual response is "GM is good at building cars. We have no experience building a ride sharing company. The probability of success is much higher if we just buy a company that is already successful."
Someone needs to wash and clean out the cars between the rides, maintain them, park them for the night as well as provide occasional repairs (i.e. have a few spare parts and bumpers on hand).
100k is the most generous interpretation of "6 figures". Would ride sharing companies still buy them at 350k each? At that point it may take 5 years to break even compared to having a human driving a 35k car
Ride sharing companies probably won't break even on the first few thousand self-driving cars, but they don't currently break even with humans either.
But once the tech is figured out, then you can start optimizing the marginal cost of each car, and spreading out capital costs over more cars, both of which reduce the cost.
What's your math for that? It really all depends on the financing deals and Uber/Lyft seem to have no problem getting money at favorable terms.
It would create a huge boost to revenue since drivers get ~75% with Uber. If they drove the same schedule it would be 4x the revenue, but autonomous vehicles don't have to sleep or see their families so the revenue difference in switching 1 human into 1 robot should be at least 6x.
False. Gross Driver Payout in most US cities is ~45% (Leaked financials)
This is mainly because:
- the booking fee goes 100% to Uber, which can be another 25% on short trips
- UberPool arbitrates rider fares and driver are paid the lower of (75% of per mile, per min, OR 75% of rider quoted fares). This accounting hack, on average, increases Uber's take to 35%
- Upfront Fares is weighed to give Uber, on average, an extra 5% to compensate for "variance".
Why would revenue surge? You assuming that rides will dramatically increase? I think you meant profit, but considering that Uber/Lyft combined are losing billions a year (Uber alone I think was over -$4B in 2016), I'm not sure a surge will be noticed.
Uber likes to keep the illusion that they aren't employing drivers, part of this means they don't record the full fare as revenue. Uber's revenue is just their take of the ride. The full amount of money is "gross bookings" which probably won't change a ton at first, but money actually going to Uber (revenue) will definitely surge.
> However, gross bookings is not the same as net revenue, which came in at $663.2 million in the first half of 2015, compared to $495.3 million in all of 2014. That's the amount Uber actually receives before costs but after it pays drivers their cut and accounts for promotions, driver incentives, and more. For example, Uber paid out $2.72 billion to driver contractors in the first half of 2015, just under 75% of bookings.
So in the first half of 2015 they had $3.63b in gross bookings, but just $663.2m in revenue. If you magically switched out all the drivers to robots revenue would match bookings which is a 5.5x boost.
If those numbers are true, I wonder how Uber list $4B in 2016? I guess I wrongly assumed they were basing revenue on actual gross bookings (which makes the losses more understandable).
Even if they are losing money, it is still a great deal - being the first to offer fully automated self driving ride sharing is going to be a HUGE competitive advantage, even if the cars are too expensive to be profitable right now.
Probably, people are expensive. Minimum wage + employer taxes for a 25 year old in the UK is roughly £78,000 for 24 hour coverage assuming no overlap and no illness, no hiring cost, no management.
I'd be interested to see what TCO they'd need to have to no longer be cheaper.
Uber doesn't necessarily need 24-hour coverage though, because demand peaks during commute hours.
However, self-driving vehicles can end up being a 'base load' that runs 24/7, and then Uber just pays drivers to appear during peak times, so they can avoid paying for (labor + driver's car) during unnecessary times.
And as the cars become cheaper, the trade-off shifts until it's easier and cheaper to have self driving cars at all times. I think there's a reasonable level of demand that would mean that even at several hundred thousand pounds/dollars, they'd still sell quite well. Every £50000 is an absolute maximum of 710 eight hour shifts.
It's a fairly recent addition, announced at the same time as a cut to in-work benefits that were mostly targeted at over 25s and a cut to corporation tax.
>"If you assume the cost of these autonomous vehicles, the very early ones, will be six figures, there aren’t very many retail customers that are willing to go out and spend that kind of money," Ableson said.
"Tesla Narrowly Misses 80,000-Vehicle Sales Goal in 2016"
Yeah I think they missed the "..on a GM" portion of that sentiment. There are plenty of people going and buying six figure cars, clearly. However, spending six figures on the Chevy brand would feel weird.
None of those 320k cars GM sold have autopilot hardware sending data back to the manufacturer for self-driving algorithm refinement. All of those Tesla vehicles do (Autopilot 2 hardware). New Teslas are delivered every day that accelerate the rate at which Tesla collects autopilot data.
Tesla fans seem to believe this very very strongly. Do you get a massive mobile bill from your car? Does your car have a huge storage device in the trunk? Does it upload 50GB of data over your wifi every time you park it in your garage? If not, the amount of data that Tesla collects for this purpose has probably been dramatically overstated.
They're not necessarily uploading continuous video of the entire time you're driving to the mothership - that would indeed be wasteful. More likely, they're monitoring diagnostic values from their autopilot algorithm (like "error in predicted steering angle") and using that to trigger recording of video and sensor values for a short period, which could then be used as "test cases" for future algorithm updates.
I think you're missing the point that a car manufacturer does not care about autopilot for autopilot's sake. They care about the feature in so far as it can drive sales and revenue.
Think about it this way: Companies like Google or Amazon don't launch products just because the tech is cool or innovative. They only launch products that can become $1B+ drivers of revenue for them.
When is the autopilot data "good enough" with any further improvements being marginal, i.e. how much of the data being sent back to Tesla is repetitive?
This problem lends itself to horizontal scaling very well, I'd say.
The only thing Tesla vehicles have the bandwidth and storage to send back is "oops i'm crashed".
It's possibly a way to test your implementation by comparing it to what the human does, but that doesn't apply for a million basic things like sign recognition and you can't teach machine learning with that.
Cost of these autonomous vehicles, the very early ones, will be six figures
GM paid a billion dollars for those Cruise clowns. If they build 10,000 self-driving cars, that alone is $100K per car.
Meanwhile, Volvo is quietly putting 100 customers in self-driving cars.[1] They're still way ahead on marketing, too.[2]
Volvo and Google/Waymo are the only self-driving car makers I'd trust right now.
They're the only ones who take the safety issues seriously. All the other ones testing in California have crappy disconnect statistics. (Uber refuses to test in California because they'd have to report their numbers, but reports of their bad driving indicate their system isn't very good.)
Until companies actually launch their products, I think nothing is a foregone conclusion. We don't know that Volvo's technology is just going to work flawlessly. People assumed that about Tesla and at least 1 person died (there are reports of accidents in China as well that tend to miss the US media).
I am strongly against the PR blitz that self-driving cars have been getting for years, but to judge a company's product before they have a product is a little presumptuous.
Each company has its pros and cons. Waymo certainly seems to have the current lead, but they bled so much talent it's strongly indicative of a serious internal cultural problem. Delphi has been steadily improving their technology over the years, but as more automakers build their own tech it will be hard for them to find a buyer. Argo is the new up and comer, but they haven't really announced anything other than a large amount of funding and some great leadership. Volvo seems to be doing a great job and I hope they deliver on their promises, but they haven't launched yet so who knows what will happen. Uber pursued the technology aggressively and that has led to something of a shitshow, and they've already lost a lot of talent. Tesla had the PR machine that is Elon Musk behind it and their customers love them (and many of their customers believe they have level 4 autonomy) but their track record for self-driving technology is questionable. Cruise has a great business model by working with Lyft and GM and have clearly developed quickly, but they haven't released nearly as much data as Waymo. Then there are the others that haven't really had any PR announcements and the companies that are completely in stealth mode.
I don't think anybody knows what's going to happen.
One of the few sources of objective data is the CA DMV's disconnect and crash reports.[1] There, Google/Waymo has two orders of magnitude fewer disconnects per mile than anybody else testing in California.
A certain level of paranoia in the design is required. The kind of paranoia that aviation engineers have. Volvo gets this - they have dual everything, lots of sensors, and their CEO says that if there's an accident in autonomous mode, it's Volvo's fault. Google got it while Chris Urmson was in charge. The others, not so much. Tesla ... many of their customers believe they have level 4 autonomy. Yes. At least one death is clearly attributable to customer overconfidence in the system.
One thing has become quite clear - expecting the driver to back up the self-driving system on a second by second basis does not work. If the driver can tune out, they will tune out. It's now clear, even to Tesla, that if you don't have level 4 automation, you must keep the driver's hands on the wheel.
> We don't know that Volvo's technology is just going to work flawlessly. People assumed that about Tesla and at least 1 person died //
There are two problems for autonomous vehicles that this part of your comment highlights - one is how good a driver is good enough, is it good enough if they have less accidents than a taxi-driver?
The other is managing the negative perception when the press report deaths, even if the vehicles are safer and produce less deaths than human drivers, the press will ensure the perception is negative despite what the statistics may show.
But it probably isn't a great comparison, at least not until we see what the GM cars can do and then the Teslas prove out equivalent capabilities (without the benefit of $50,000 of sensor hardware).
I also think they missed the mark on "six figures". If that car costs six figures they are DOA. The most expensive part of self-driving cars (as I understand it) is the LIDAR, and I think those have come down to sub $10,000, and Tesla has more-or-less figured out how to do self-driving without it.
I would expect that the engineering costs to bring the vehicle to market would, even when divided among many sales, far exceed the cost of a single LIDAR unit.
>"Tesla Narrowly Misses 80,000-Vehicle Sales Goal in 2016"
So 1/10 of 1% market share a whole lot of red ink (meaning cars may be under-priced, increasing quantity demanded), and no competition constitutes "many customers"? I guess it's all relative, but I disagree.
>"If you assume the cost of these autonomous vehicles, the very early ones, will be six figures, there aren’t very many retail customers that are willing to go out and spend that kind of money," Ableson said.
Oh well I guess it will be a few years before I can recline and take a nap in my robo car while in traffic
Why not? It's a technology problem and technology almost magically gets cheaper over time. I don't see why it won't eventually be like all the other required safety features (airbags, ABS, etc).
If you live in the middle of SF it might sound silly to own an autonomous vehicle, but if you live out in rural Sonoma County it is going to sound a lot less silly. I frequently go out to the mountains and I doubt I'm going to want to call an Uber to load all my camping gear into and be gone for a week.
Furthermore, if it's your own vehicle, fully autonomous mode under certain pre-defined conditions and roads can still be a huge win. But that constraint makes the same vehicles basically uninteresting as a shared vehicle taxi service for the near term.
It's great that all this work is being done but IMO summoning Johnny cab to take you home from a bar in SF is decades away.
Every Cruise Automation Chevy Bolt is using two Velodyne HDL-32E, each of which has a list price of $30,000. Price aside, Velodyne is really struggling to produce these units. Some recent HN comments have mentioned lead times of 40 weeks [1], though the upcoming Velodyne factory should alleviate these issues in 2018 [1].
We will need miracles in lidar technology for GM to hit their target of thousands of vehicles next year.
Miracles in Lidar are coming. Quanergy, Infineon, Innoviz and Velodyne are all rushing to technologize solid state Lidar, which offers comparable performance specs to the spinning kfc bucket at quoted prices ranging from $250-$40. Imagine all the places we'll be using Lidar if it gets that cheap, the applications extend well beyond self driving cars.
From a capitalism point of view, the high price and long lead times imply that the market is signalling that it would like new entrants. Does Velodyne own patents that would keep out other players?
It will probably take longer than that. I've yet to see a non-tech demo fully self-driving car. If I'm being optimistic, I'd say 2020 at the very least for early models.
I have question given that 2018 is the date to deploy test vehicles how much time do they need to their tech operationally mature?
I see in the article that they acquired a self-driving tech startup but that doesn't say much about how much experience that start up brings, and it not like there's many veterans in this field.
Will they be playing catch up with Tesla, Google et al for years to come or will the tech soon start becoming commoditized?
I agree with the folks who see this as a serious risk to TSLA. If you've driven a Model S or Model X you know that the design, the software and the performance are insane, but the fit and finish (especially of the interior) is like a 1980s Cadillac. When BMW and Benz start making EVs, TSLA is going to be pushed hard and could likely end up becoming a software or powertrain T1 supplier.
1. People are not going to avoid a tesla on the basis of a slightly crummy interior, they sell based on virtue signalling.
2. The mass produced version of the Tesla is yet to be revealed and it may have a better finish so this may be a moot point anyway.
This is more a comment on TSLA than GM, but I still feel it's relevant given the competition between the two.
I don't understand the Gigafactory, from a capital allocation point of view.
Geographically, it's not near any major lithium, cobalt, (economical to extract) rare earth, etc deposits, and it's not in a traditional manufacturing area. Maybe they're getting a sweet tax deal. There are a number of tangential benefits, like heat recycling, solar availability, USA made (given the currently protectionist administration, but that could change in less than 4 years, max 8) etc, but I'm not sure these benefits outweigh the cost of being so far from raw material and export lines.
Obviously, the Gigafactory makes batteries, a technology which improves at an unbelievably slow rate compared to virtually anything else. However, battery research is an intense area of research focus, and the whole factory is one Nobel from becoming obselete, barring (probably expensive) retooling. A game changing improvement is not a matter of if, but when.
TSLA's bet is that this does not happen until they've paid off the asset. If you makes batteries as a core business, investment in facilities is the cost of doing business. If you make cars, it's a silly and imprudent experiment in vertical integration. Why bet the farm when you can pay marginally more for your batteries and cement the luxury market you already own? Batteries are a commodity, and at TSLA's scale, they surely can negotiate a good deal. Realistically, TSLA's target market is fine with absorbing a 2 or 3% increase in battery cost that a 3rd party supplier would demand.
Given the investment in vertical integration, it seems TSLA is trying to go after the low/mid-market. Why beats me -- razor thin margins necessitate scaling incredibly quickly. They are taking an incredible risk for meager rewards in fiercely contested territory. Contrary to what most outlets will tell you, Ford/GM/Japanese automakers are not stupid. Slow, yes, but not stupid. They will move quickly to shred TSLA in the low end market, assuming they recognize the threat. Not only that, but they take on comparatively little additional risk in doing so, outside of product development costs. They don't need to build assembly lines, factories etc, because they already exist. Even if one mfg misses the electric vehicle trend, it's highly unlikely 6 of them will.
That's all, and I'd really like a counterargument.
I heard Elon say that if they were to produce 1M cars per year, they would need to consume all Lithium-Ion batteries produced in the year 2014. Obviously there will be more batteries produced in 2020, but Tesla's fate would be at the mercy of the battery producers, which could collude to drive the price of batteries.
For better or worse, Tesla is hell-bent on vertically integrating every single aspect of production. I believe they might be working on software that will allow them to stop relying on NVDA's chips for self driving as well. "First principles" business model will make or break Tesla/Elon.
The answer is pretty simple; global production of battery cells suitable for automotive use is far, far below what is required for Tesla's goal of turning (forcing) the majority of the world's produced cars electric. Buying this capacity from a third party cannot be done; there is no seller capable of providing the product.
Global car production is on the order of 50 million vehicles per year, which means a required capacity of 2500GWh/year assuming 50GWh/car. 2013 li-ion battery production was 35GWh, some 1.5% of this number.
Tesla wants to reach a production rate of 500k cars/year in the next couple of years, and you probably wouldn't find a 3rd party manufacturer willing to eat the risk of upping the world's battery capacity by 50% to fulfil this contract. Never mind Tesla's long-term production goals, which are in the millions of vehicles per year.
Hence they're forced to do it themselves, same situation as the charger network on a bigger scale (which could in an ideal world be done by governments or a third party, but never on the short time scale required). I can't speak to the location of the factory.
> Geographically, it's not near any major lithium, cobalt, (economical to extract) rare earth, etc deposits, and it's not in a traditional manufacturing area. Maybe they're getting a sweet tax deal
Land is cheap, labor is cheap, and it's on a major rail line. You don't need to extract the lithium or cobalt on site, you just need to be able to get it cheaply. If this [0] source plays out, then they actually do have lithium nearby.
"Bolt"? They really couldn't name it something more identifying and clever than the self-driving bolt?
I came here thinking they were attempting some way to make things self assemble. "It's a self driving bolt. You don't have to use a screw driver to screw it in it does that itself."
Besides, GM building a self driving car feels like letting the people who make plastic grocery bags make doors for safes or something. Sounds dangerous and it just sounds like it'll cause the name "Bolt" to stand for "run away suddenly out of control."...
I kind of covered my personal answer to that in my original comment.
It also annoys me that companies name things common English names because that makes it hard to search for things. "Chevy bolt loose steering" "bolt bolt size trunk latch" "bolt size for tires"
Ultimately, it's obviously a battle that I'm not going to win because I'd prefer everything just have a GUID so I can easily identify exactly what I have. "The Chevrolet B0F30517-BAFE-43EB-BA62-1832AC5272F7"
Lastly, I know it's been a common theme for a long time but it just seems to keep getting worse. Apple Ping. Chevy Bolt. Chevy Volt. Chevy Spark. Many different products called Hydrogen etc. Rust. I know of at least 2 products called Go. It just wouldn't be difficult to change spelling or do something clever instead of just taking direct, common dictionary words and grabbing them for your product.
Search engines easily handle things like [Chevy bolt] once they become popular enough. [chevy bolt size] in google already shows me stuff related to the Bolt.
> Besides, GM building a self driving car feels like letting the people who make plastic grocery bags make doors for safes or something.
While it's true that GM isn't exactly known for their construction quality, they seemed to have somewhat turned that around recently, and have definitely embraced making good software.
Also, their self driving tech comes from the acquisition of Cruise (a YC company).
- Incumbent auto giants entrance diminishing Tesla's margins.
- Incumbent expertise and brand that diminishes Tesla's appeal.
- First mover disadvantage: rest of auto industry sees large initial ROI by piggybacking off Tesla
GM can't be the only guys racing to bring Tesla experience to it's existing brand followers. Luxury makers like BMW, Benz will woo the lion share of the market Tesla originally pioneered. First movers rarely do well here unless they've cemented their brand position in the luxury space such as Ferrari, Lamborghini etc.
tl;dr: When the status quo catches up to Tesla, it will put downward pressure. Tesla's winning strategy should be to limit quantity and artificially inflate unit price. It is not equipped to take on dozen automotive brands in the long run.
Please correct me where I'm wrong.