Tesla's Next-Generation Model Y Chassis | Hacker News Reader

Tesla's Next-Generation Model Y Chassis

blog.onlyusedtesla.com

113 pointssimonebrunozzi161 comments

Open articleView on HN

> Tesla has begun building the next generation of Model Y with singular front and rear body castings.

This is not accurate.

They are using castings for the front and rear chassis along with the structural battery.

Bodies of Tesla cars will continue to be stamped, which makes sense because bodies are mostly cosmetic sheetmetal.

https://www.sae.org/news/2020/06/tesla-model-y-big-castings

> Munro maintained. And “castings don’t repair very well,” he added. “If an impact was severe enough, the car’s a write-off,” he said.

I own a Tesla Model 3.

In the past I owned a GM vehicle which was damaged in a collision. The chassis had a few parts which were bent. The minute a chassis of any modern vehicle is bent in any way it is an automatic write-off in the insurance industry. This is exactly what I was told when my car was categorized as totaled.

It doesn't matter if the chassis is composed of many parts and can theoretically be repaired. It is such a huge undertaking to disassemble the vehicle to a level where you can repair the chassis that nobody in their right mind wants to deal with it or pay for it.

This isn’t a given. It depends on what a repair entails. Often a steel chassis can simply be pulled back into shape.

An aluminum chassis is certainly more difficult. A friend of mine had a Lotus and the whole thing was written off for chassis damage that literally couldn’t be seen.

How was it known to be damaged if you couldn’t see it?

I believe lasers are used to check how various parts of the car align. Certain things out of alignment indicate a bent frame.

Could see chassis damage underneath. The body looked perfect.

Do any states that allow bent frame repair without a salvage title? It's been a while, but back when I was buying cars at wholesale auctions, it was a big deal when one slipped through with a bent frame without a salvage title. They reversed auctions all the time, including one truck we purchased.

If it's repaired and not totaled, it won't have a salvage title.

I have even "bought back" cars that my insurance wanted to write off (insurance paid me market value less their estimated salvage value) and there was no change in title.

I think the only time a title gets "branded" is if there is a change in ownership. So if the car is totaled/written off, and sold on to a salvage yard or salvage auction, and then repaired and sold again, it will have a "rebuilt" title.

Yes, Virginia does. Happened to my Jeep a few years back.

A steel frame can be pulled into shape. A steel unibody is another can of worms. It'll never be straight.

I had a WRX pulled back into shape.

Maybe we’re talking about different things? Maybe there’s some level of unibody damage that crosses the threshold? Because some damage to the unibody is pretty much guaranteed in any moderate accident. The whole vehicle essentially is the unibody.

Yes, bent unibodies can often be straightened by a body shop with the proper equipment. Of course serious damage where the body is really crushed or severely bent or broken might not be economical to repair.

And as a corollary, I knew someone who had his Porsche sandwiched in a sudden traffic stop and it needed about $90,000 of repairs, including the chassis being unwarped, as I understand it, and the insurance paid for the repairs because it was marginally (~10%) cheaper than replacing the car.

Absolutely false. I had a new Jeep Wrangler get rear-ended (by a Smart Car of all things) with $13k+ in damage, including a bent frame, bent floor tub, and twisted roll cage, all of which was repaired.

Wrangler is body on frame.

The minute a chassis of any modern vehicle is bent in any way it is an automatic write-off

So…? The statement to which I responded didn’t make any distinction and is wrong regardless. Unibodies and body-on-frame can both be straightened. The former is more difficult, but doable in many accidents.

I’m guessing body on frame does not qualify as “modern”. What was the last passenger vehicle as opposed to truck built body on frame? The Crown Vic beloved of police departments?

The comment you're responding to is making a distinction between the body and chassis. The Model Y has a cast chassis, but the blog article says it has "singular front and rear body castings".

> The minute a chassis of any modern vehicle is bent in any way it is an automatic write-off in the insurance industry.

This is true much more often than it was in the past, but it’s certainly not always the case.

I think it will be cool if they found a cost-effective way to make vehicles more modular like body on frame vehicles are, but without the tradeoffs.

There’s not enough detail about the process and materials here, but one of the issues with castings generally is that they sometimes fracture rather than bend.

Anyway, maybe this is a way to bring the cost of new vehicles down for more affordable models in the future.

Possible with the modular skateboard model:

https://mondaynote.com/evs-the-manufacturing-revolution-40a7...

And even if it was cost-effective to "unbend" the frame elements, they wouldn't be as strong as before because plastic deformation changes the crystal structure of the metal.

What is “plastic” deformation in the context of metal?

"Elastic" deformation means the material deforms under load and then returns to its original shape after the load is removed.

"Plastic" deformation means after the load is removed, the material does not return to its original shape.

A spring is the canonical example of elastic deformation; it returns to its original position when the load is removed. A dent in a car (or a bent frame) is an example of plastic deformation.

The hydrocarbon compound we know of as "plastic" got its name from this phenomenon (more precisely both the compound and the phenomenon got their names from a Greek origin word meaning "capable of being shaped or molded".)

https://en.wikipedia.org/wiki/Plasticity_(physics)

https://en.wikipedia.org/wiki/Plastic#Etymology

you do not straighten, you cut out and weld replacement pieces

Nice, but no more body and frame repairs. One dent to those big parts and it's time for a new car.

As long as insurance premiums stay reasonable, it’s a win for manufacturing speed and complexity, product quality, as well as the salvage market. And in my personal experience, vehicles are never quite the same after frame damage repairs.

(legacy [Aug 2021 build, hah] Y owner)

> And in my personal experience, vehicles are never quite the same after frame damage repairs.

Same, a car with a damaged and repaired frame is usually a different car than the original. Overall this is a win.

Fwiw this isn't the first mass-produced consumer vehicle with a unibody construction. For one example, the Nissan Pathfinder had it back in the 90s, though later returned to ladder-on-frame.

https://en.wikipedia.org/wiki/Nissan_Pathfinder#Second_gener...

> this isn't the first mass-produced consumer vehicle with a unibody construction

Isn't it already unibody construction, along with basically every other non-truck vehicle in production (and some trucks, too)? The only thing new is using casting for so much of the structure.

The unibody construction isn’t unique at all in 2022. Virtually every car on the road (except for handful of pickup and some SUV/off-roader models) has been built this way for decades now, and isn’t what’s exciting about this new approach from Tesla.

What’s unique is that so much of the unibody is just 2 castings - previously the front and rear required joining 140 separate parts. Such a massive front and rear casting has never been attempted in car production, the mechanical press used to do so is a record size for a car company.

VW’s ceo Herbert deiss has stated this is a massive innovation from Tesla that ultimately will enormously reduce time to assemble a vehicle - we don’t need to just take elons word for it. Time to assemble is everything when it comes to car profits. Herbert points out a VW ID3 takes ~3 days to finish assembly, with this process Tesla should be able to complete a model Y in just 10 hours, putting large competitive pressure on VW too. VW’s margins on their electric stuff hasn’t been that great to begin with either.

Hopefully these huge casts can be recycled and recast. If so, they don’t seem terribly wasteful. I’d also hope that minor repair of the cast part is possible for the smallest dents. I’m not an expert in this area.

Metal is extremely recyclable

Especially aluminum which I believe is what Model 3 and Ys are made out of

No, you're thinking of the Model S. The 3 and Y are predominantly steel.

Ah, I have it backwards.

The large chassis castings in the Model Y, however, are an aluminum alloy.

I wonder if having less structural adhesive between metal components makes it more easily recyclable.

We'll see how things pan out, but this might actually make repairs a better experience over time - fewer types of parts means it's easier to keep them in stock, and likely less time required per replacement.

Tesla can't keep parts in stock because they're desperate to prioritize production over parts inventory, not because they have too many parts.

This will be a nightmare. A damaged fender now means instead of a fender-sized package that easily fits in a truck with lots of other things, you now have a massive half-a-car sized package.

This also doesn't bode well for parts availability in the future. I can't see Tesla switching these massive presses back to producing older designs of panels.

Why would a fender replacement require this? The castings are for the chassis, not the body.

I don't think so. I think it will mean more write offs. These castings replace current pressed panels, so anything in that area will now mean a new body shell -which realistically will mean a written off vehicle - where repairs would currently be an option.

I think this is manufacturing optimisation at the cost of overall life optimisation, like the epoxy-encased batteries which are less recyclable than the previous generation.

Other commenters have pointed it out in different threads, but these castings won't replace the panels. They will replace the current chassis.

I'm aware of that. But using those terms is somewhat misleading - modern vehicles use a monocoque chassis, and these castings will be in areas which are routinely damaged in collisions where panels can currently be replaced/repaired by any competent bodyshop. You can usually buy such sections individually, unpick the originals and replace with new ones, returning the vehicle to as-new condition - even when you're a DIY mechanic with suitable equipment.

That won't be the case with castings, and even less so with Tesla given their approach to third-party repairs.

I really doubt that replacing one of these cast parts is feasible -- everything is attached to them, you'd have to take apart and rebuild almost the whole car.

As far as I remember from my material sciences classes, casting is a lot more difficult to model and QC than stamped and welded/glued steel panels. This is also a completely new process for automotive at this scale which complicates it further.

I am sure they hired the best engineers and tested as best as they can, but this seems so new that I expect there to be a lot of problems with the first year or two of production -- maybe a reason why this process is coming to North America last. And given how castings usually fail, I don't expect failures to happen immediately, but after 3-5 years or more.

But then again, I'd have said the same about using "laptop batteries" for electric cars.

> maybe a reason why this process is coming to North America last.

Fyi, the article says it's coming to Gigafactory Berlin and Gigafactory Texas first. It's only coming to Gigafactory Freemont last.

Doesn't the Y from Fremont already have the rear casting, as of January 2021?

> This is also a completely new process for automotive at this scale which complicates it further.

The Trabant had supply limitations throughout its entire production run due to scaling issues casting Its duraplast body.

The model Y, as delivered today, already has a very large casting in the rear: There are many sources out there describing it, and how it simplifies model Y assembly compared to the otherwise rather similar model 3. What they are describing here is to go further with a process they are already doing very advanced work on in production cars. They might have problems. They can fail altogether. Maybe it turns out to be a bad idea anyway, but I do not expect Tesla engineering to be just naive at this.

As for why Fremont is the last place to get the changes, it seems to me that there are simpler reasons: That factory is the main source for Teslas on the road today, and it operates more or less at capacity. Adding a new process would slow it down, at least temporarily. The Berlin and Texas factories, AFAIK, have yet to start operation, so they are good places to try something new, given that Tesla is selling the cars about as fast as they can make them. This is how a lot of the industry does things: Make your new facilities have the latest and greatest, and only retrofit the old ones when it makes economic sense to dismantle the older process. No matter what the theoretical savings might be for a new casting vs so much soldering and assembly, the old process is making cars, today, so decommissioning it is not as good a value proposition as just not even trying the old process in a new facility.

They will be designing the cars with buffers, but if a crash can go through all of that AND damages the single casting chassis, which are a lot stronger, then you should thankful because you'll likely have gotten way more injured with a traditional chassis.

Damage in a collision is a good thing as it requires energy. Not sure why you say a stronger part will reduce injuries.

Perhaps because the stronger part takes more energy to damage and thus absorbs more?

Absorbs? What does it do with that energy, then? Radiate it? AFAIK it's just stiffer so it will effectively transmit the energy along to the next part in the chain. Eventually, that'll be a human.

I meant the energy will be used to deform the chassis, but perhaps I am over my skis here.

Into heat mostly. It takes quite a bit of energy to deform a lump of metal. You need that energy for force atoms past each other, and the metal gets hot in the process.

So yes, absorbs the energy, then radiates it as heat.

The point that's being made is-- if there's a certain amount of energy being dissipated in a collision, it's better for things to crumple so that the energy is dissipated over a long time period and the peak accelerations (which damage occupants) are less.

A hypothetical perfectly rigid car that stops instantly in a collision instead of deforming is not safer for the occupants, because suddenly they will be going 40MPH inside a not-moving car.

Because they’re saying these cast are the inner shell that forms the passenger compartment. If they’re deforming, then they’re deforming into the space where the passengers are, and humans squish much easier that aluminium castings.

Regardless if it’s a unibody or a welded chassis parts of it would be designed to give under stress and parts of it would be designed to hold to protect the driver.

In a bad accident with any BEV i would be far more worried about the battery being compromised than the chassis tho.

Aren't body and frame repairs for Teslas already pretty bad? I've looked at the insurance rates for used Model S and they're massive.

I pay less than $100/month in insurance for a 3 year old Model S. Our Model Y is about $130/month. Sunbelt state, top tier insurance company.

I pay about 900€/year for my Mercedes and a Model S would've been something ridiculous like 1800€/year.

I presume that you are also in Germany as well. The insurance rates here are straight up highway robbery. I wanted to get a Tesla, but the insurance quoted me more than 2000 euros.

Nah, Finland. But I guess there's some consistency in Tesla insurance rates being higher than other cars, so I guess that they're more prone to get into crashes, and/or they're massively expensive to repair.

The outer body of the car is still sheet metal and can be repaired as before, if not easier. The cast parts are only on the inside and also traditional cars are usuall a write-off, if the internal frame is damaged. The cast parts might even be stiffer and hence more difficult to damage.

Tesla for sure has thought about this. For example, the rear cast has struts which hold the rear bumper and consequently are likely to be damaged during an accident. But these struts are designed to be cut off and replaced with new parts.

> The cast parts might even be stiffer and hence more difficult to damage.

Someone with more experience should step in and correct me but my understanding is that this is how we used to do things and stopped because it turns your passengers into soup. You need those parts to crumple. I'm sure Tesla has thought of that and wouldn't downgrade the structural safety but it does mean the new design is probably just as easy to damage but more expensive to replace than traditional designs.

Cars are designed with a strong central structure (passenger cell) and weak areas at the periphery (crumple zones). These cast parts are in the "strong" area, I think.

I suppose you want the crumble zone to be on the outside. Outside critical parts and easily replaceable. Then you want the inner shell not to disform, or not too much, as otherwise you are sending metal into body parts.

The crumple zones are unaffected by the casts and the casts can crumble as well. But if the car gets deformed beyond a certain point, it would be totalled anyway. If you look at pictures of the casts, the front ones mount the wheels but end there, the rest of the car ahead of them is made from sheet metal and metal bar with the purpose of crumbling easily.

A steel unibody may bend in a moderate collision, but it can often be straigtened by a competent body shop. Castings are more brittle, they tend to crack, not bend. And they can't really be welded back together. I think having major parts of the frame as cast pieces will reduce repairability, but time will tell.

If you want to read more about the actual “Giga Press” see https://en.m.wikipedia.org/wiki/Giga_Press

It’s made by Idra Group in Italy, it’s only a matter of time until other car manufactures also adopt it.

The group seems to be a relatively small Company, and have sold only 19 such machines around the world. There must be other manufacturers producing this sort of tech, I guess?

Welcome to Europe where small companies you never heard of make hyper-specialized machines and tools that are being used around the world to manufacture your household items.

People way underrate how much sophisticated work gets done in Europe. Great comment

Another example is https://en.wikipedia.org/wiki/ASML_Holding

Or https://www.krones.com/en/company.php

Makes machines for packaging products

Not really, the incumbent US manufacturers have too much invested in traditional manufacturing techniques to switch over. The new Chinese automakers are the ones that will be implementing it next.

>> Not really, the incumbent US manufacturers have too much invested in traditional manufacturing techniques to switch over.

Sunk cost fallacy. Yes, they tend to fall for that. Sometimes they get past it.

Some of the incumbents have existed for more than a century, saying that they're not capable of changing manufacturing technique is quite silly.

https://www.autoevolution.com/news/bmw-s-ceo-says-tesla-s-ma...

>there is no economic reason to have very large integrated parts in these vehicles.

Thank you for the article. It says that BMW doesn't economic reason, but Audi and Mercedes are adopting it.

The CEO of BMW saying they’re concerned with repair costs is a joke.

Big presses/forging-machines are perhaps an underappreciated asset. They open up otherwise-impossible possibilities.

https://en.wikipedia.org/wiki/Heavy_Press_Program

Pretty epic feat of engineering! If this technology spreads I wonder what will happen in countries like Germany where car manufacturing employs a huge swath of the population.

Already with electric cars we have way fewer moving parts, now also the assembly line can be drastically reduced. I am curious to see what will happen in such places.

Instead of controlling welding robots, they control stamping machines?

German mfgs are already investing heavily in automation.

What percentage of Germany's population actually work in car manufacturing? A quick google search reveals 300k.. and Germany has has a population of 83 million.

For example, BMW production exists in five different countries. You can find BMW manufacturing facilities in Germany, Mexico, China, South Africa, and America. The Spartanburg plant in Greer, South Carolina is the largest BMW production facility they have (by production volume).

It's way more than 300,000. Maybe people physically touching cars and car parts but there are thousands of small companies making very small parts like lighting, the handle for seat adjustments, the foam that goes into the seat, small plastic parts, and many more.

Numbers I could find in a study made by the research group of the German parliament [1] were 800,000 and those are only people directly employed by manufacturers and their suppliers.

1: https://www.bundestag.de/resource/blob/496346/1a9b4fbe228b43... (German)

IIRC the legacy automakers are already heavily invested in automation. I would not be surprised if Tesla's production lines are in fact a good bit less automated than, say, BMW's.

From quality reviews I have heard that TSLA has more deviation in things such as gaps in panels which are unheard w/ traditional automakers...so I'd assume less automation would result in things like this.

It's not just hearsay. I've driven a Tesla and seen a bunch of used ones up for sale. The gaps are hilariously huge on one side, and too small to allow a finger nail in on the other side.

Each car varies in their gap consistency, with some "good ones" and some lemons. This inconsistency alone should be a cause for concern.

I spent 20+ years working as an automation consultant in automotive manufacturing. All the technology that Tesla has available to them, is being used by the other manufacturers as well. Tesla sometimes is less cautious on when they deploy that technology but that’s about it.

From a manufacturing standpoint, this will only impact stamping/casting and BIW (body in white) operations. Both are highly automated already with very few people (think the robotic lines you would see in a movie). Casting just takes work about from the stamping press and welding/riveting robots.

As a last point, from the plants I’ve been in (including Tesla), Tesla is actuality LESS automated than their counterparts in a lot of areas, because they are willing to “throw labor” at a problem at the expense of cost and quality in order to get parts made. You can do things way less efficient when the stock markets doesn’t care about your bottom line profits.

I‘ve often heard that Tesla had to built much of their automation machinery themselves because there were no suppliers that could build „the machine that builds the machine“. This seems contradictory to your statement, do you think there is no truth to this kind of stories?

Edit: I’m also wondering why Tesla is supposedly able to produce a car in 10 hours vs. about 30 ours for the traditional manufacturers? E.g. https://www.torquenews.com/14335/elon-musk-manufacturing-wil...

The only thing different between Tesla’s and GM/Toyota/Volkswagens manufacturing processes is the marketing. They all use the same equipment.

Most manufacturing lines (robots, presses, equipment) etc are built by dedicated machine builders (not unlike ASML for chip fabs). This is a very mature industry.

GM, Volkswagen, Toyota, etc, would build 8M vehicles a year, across dozens of products lines, and refresh their vehicles ever 3-4 years. Tesla just approached 1M last year. Tesla is not a major customer in this space yet, so in some cases they just bought a small machine builder to make there lines for them. (Although granted equipment manufacturers do love to list’s them as a customer for the “cool” factor).

An electric car is way simpler to manufacture. Think of how many moving parts are in an engine and transmission alone. So many in fact that engine plants and transmission plants are there own separate 3M+ sq ft facilities. I would think the decreased manufacturing time would come down to this, because from what I have seen Tesla plants run way less efficient than any of the established auto manufacturers.

Indeed, what other manufacturers do not have is a Twitter account the size of @elonmusk

This is really cool and super interesting. Maybe someone who knows more about cars can answer: does using a single piece of metal make the car harder to repair in case of a collision? Does it make it more likely to consider the car “totaled” in case of a collision? Is that good or bad for Tesla and/or customers?

With welded parts you can just drill spotwelds out and replace the part, with single piece, you cant even get new part because they only make whole piece. If you manage to get parts, fix wont be as good as new because now there is a weld when it was solid part before.

This is more of a broader question, but when is it a good time to buy an EV? As someone who just recently got a model Y, seeing reports like this just frustrate me. I thought about waiting until the rumored refresh but decided to go for it. Why get an EV if a newer one that's better in almost every way (range, cost, etc) is constantly just around the corner.

As an example, I'm not kicking myself when the "all-new redesigned" 2021 Hyundai Tuscon comes out because while the face lift is nice, the car isn't that much better than the 2020 version. It gets similar gas mileage and has similar features. Sure there's a few nice things on the newer Tuscon but half of them could be on the old one with a software update.

You have to decouple the purchase from the potential future, just like we used to do in the late 90s/early 2000s with computers.

So when it comes time, evaluate what is available and make a decision then - it's unlikely something comes out that makes all previous ones stop working after all.

And you do get the utility during that time.

If you're really concerned, look into leasing; they usually have a buyout option at the end.

Am I the only one who doesn't want to buy "all-new redesigned" cars. I won't want an all new redesigned Operating System.

Why? There are far fewer parts in a car than an OS.

That doesn't really track since cars themselves have multiple operating systems.

I don't want a mandatory software update that changes where my rear window defroster button is. I want a button. Is it so much to ask for a 90s BMW interior with a Tesla drive train? I think that's what GP is getting at.

Modern cars are rolling local area networks that use a fragile real-time network (CAN bus) in addition to 4G bluetooth, and bespoke wireless network protocols.

They include many operating systems, and all sorts of other unfathomable complexity.

Many cars as far back as 2003, and all cars model year 2008 or newer in the US, use CAN Bus. Moreover, as far as real-time networking goes, CAN is pretty far from fragile.

I agree that it is more reliable than most realtime network protocols, but that’s a pretty low bar.

Two of the three cars I’ve had with CAN busses had recalls due to CAN priority inversions. One couldn’t turn the ABS off when the rest of the car shutdown, so it would sometimes drain the battery if parked for more than 6 hours.

The newer one gets into a state where the “turn off cruise control” signal gets preempted. The solution is to stand on the brake pedal and force it into neutral. (The transmission control is also on the CAN bus, from what I can tell.). Those two events together put the power train into some sort of panic mode, and it stops applying throttle to the engine.

The fact that the network protocol has been around for a while is no reason to trust a first-year redesign of a vehicle platform.

No. Redesigned cars are less reliable.

https://www.consumerreports.org/car-reliability-owner-satisf...

> CR’s proprietary analysis shows that vehicles tend to be most reliable by the final year of any particular model run (typically five to seven years), after many of the bugs have been worked out, and least reliable in the first year of a redesign, when freshly reconfigured and often touted as “all new.”

> Why get an EV if a newer one that's better in almost every way (range, cost, etc) is constantly just around the corner.

It's just the nature of the beast when you're looking at a rapidly developing technology. Think of it like buying a cell phone in 2008.

Buy it when it makes sense for you as it exists, and accept that anything you buy today is going to seem like old news in a few years.

That said, the real developments as I see it have been in battery tech and fast charging. Unless you really care about the infotainment system or nonsense claims of the vehicle driving itself I can't say that I consider really any 250+ mile fast charge capable EV to be "obsolete". A first-gen Model S is still a solid EV almost a decade later.

Two thoughts.

1) The biggest benefit of this is manufacturing speed and cost neither of which affect you, especially since prices have been rising due to demand outstripping supply at the moment would be kind of silly to put off buying a car you other wise need to get this.

2) Despite what this article implies your Y’s rear end is already built with casting construction. The difference in Texas is they are also going to start doing it for the front piece. But again it is not something the end user would notice so unless you are trying to calculate Tesla profit margins it kind of doesn’t matter.

A good time to buy an EV is when you can afford one and it fits your needs. I don't regret buying my first gen i3 years ago. I also got a model y recently and it's a much better ev but the i3 is still useful.

Tesla is one of those cars which gets BETTER with time. I have a 2018 Model 3, it is a MUCH better car than it was 3 years back. Not just cosmetic changes, they UPGRADED my CPU to handle more advance calculations (for free), constant software updates to make it better every 2 weeks. I reminds me of iOS updates, but more frequent & more consequential (added features which let me monitor the parking lot remotely/visually before reaching the car etc)

Our Battery packs are good for 400,000-1million miles depending on model. What more could u ask for? 420,000 mile? 1.1m miles? The point is, it is already as good as it gets as of TODAY, future efficiencies will always be there. But that is out in the future. Also, these are MANUFACTURING efficiencies , the end result for consumer is same/almost same.

I was an exclusive BMW driver for 15 years before moved to Model 3 - I enjoy every bit of it. Never ever had buyers remorse.

> they UPGRADED my CPU to handle more advance calculations (for free)

No they didn’t, I assume you are taking about the Autopilot computer upgrade, that upgrade isn’t free, you only get it if you’ve paid for the FSD upgrade and the are only doing it because the previous computer wasn’t powerful enough to deliver FSD.

yes - but I originally had paid for FSD half of what it costs today ($12k). They could have easily asked me to pay the difference.

It comes with a premiere customer experience, not a nickel n dime car dealership experience.

So you paid 12k for a feature that was delivered to you 3 years after getting your car?

Tell us more.

To me, the decision would be driven by access to charge networks.

If you road trip a bunch, it definitely makes a difference. It was less hassle to road trip my Model 3 than it is to road trip my wife's Bolt.

However, that said, for every 1 time I use a DC fast charger, I charge up at home perhaps 150 times. And that experience is the same for any EV. For one or two mild road trips a year to visit relatives, it isn't a huge factor in which car I choose; the day-to-day experience is much more important.

Right. I think people overwhelmingly recharge at home, and if you live in, say, an NYC or Chicago apartment building, owning an EV can be a major hassle. I have almost never seen Teslas parked in Manhattan streets, but in outer boroughs, private residences have a Tesla at every 3rd house (I am barely exaggerating).

Number one rule of buying a Tesla: be happy with what you got when you got it for how much you paid. They redesign then continuously, so there's never a "good" time to buy.

For example, over the next 12-18 months the Model Y will start getting front castings (but only ones coming out from Texas), AMD Ryzen, then 2170 cells, then FSD hardware v4. But if you decide to wait for these things, then there'll be another slew of changes waiting to happen.

I dunno.

The marginal value to waiting another year is likely much higher today than it will be a few years from now. The newer tech is, the more value there is to waiting generally.

Of course, really just depends on your personality, and how cost constrained you are.

I specifically chose not to buy an EV because I know within 5 years the tech will be much better, and infrastructure won't be as much of a pain.

That's 5 years where you won't have an EV - depends on what the utility you get in those 5 years.

You're right that utility and opportunity cost of that lost utility matters.

My personal utility for an EV is largely negative in the short term though. Too bleeding edge, higher risk that technology shifts such that current models devalue quickly, fewer charging stations than gas stations. Current range is sufficient at purchase, but how does range change as vehicle ages?

A lot of those things will be much improved over the coming years. I don't gain anything from converting in the short term other than the aspirational/goodwill parts

Consider that Tesla has been consistently raising prices due to parts and demand, the best time was probably yesterday.

You just have classic FOMO and with Teslas, you have to override your monkey brain to stop wanting the latest and greatest. When I was younger in my 20s, my brain compelled me to buy the newest Android. I loved having the new tech and I always thought I could sell my old phones before they became obsolete.

Teslas feel very similar to that high you get from having the latest phone. I own a 2017 model S and it seems every 6 months my car becomes more legacy.

First it was all the packages (Cold weather, premium upgrades, etc) becoming standard when you had to pay for the packages.

Then it was HW 3.0 & MCU 2 for advanced screen graphics and playing Netflix. Also capability for Full Self driving.

Then it was the raven air suspension.

Then it was model 3 and model Ys getting external pedestrian speakers that could play music.

Then it was a heat pump.

Then it was being to roll your windows up and down through the phone app.

Then it was faster super charging.

Then it was new, fun video games.

Then it was a rectangular screen that can play witcher and tilt (eventually)

Then it was more battery range.

...and so on.

That's all I could list off my memory but I bet you I missed 50% of the updates. The moment you buy a Tesla, it'll be out of date in a couple of months. It is what it is.

Came here to look for this list ;-)

Don't get me started on my new Model Y that doesn't have the game controller-only games... or my Model S that doesn't let me actually play games in the back. Wait, it's a car, I'll be fine.

It's almost best to disconnect from the forums and fans after taking delivery.

The ordering, waiting, and delivery process really do require us all to connect - if only for mutual support through frustration -

There is an aspect of "who cares" when I hear about slightly different door panels, or heated windshield wipers, or a manufacturing improvement for newer cars being built.

Yet it's easy to get lost in the "OMG THE NEW HEADLIGHTS MY CAR IS RUBBISH" world, too.

Enjoy any EV, when you're ready, it'll be a great car for many years.

I leased my Bolt. I figured I would buy it out when the lease expires unless current EVs are competitive, but now, I don't know, maybe I'll just keep leasing. EVs feel less like buying a car and more like buying a really expensive PC in that the minute you buy it, it already seems somewhat outdated.

Sure, you can get that feeling with an IC car, but IME the updates are less frequent and the technology is more mature. My '17 Subaru feels a bit dated, but that's nothing I couldn't fix with a simple head unit upgrade.

When leasing, do you have to be extra-careful with not dirtying/brushing/hitting the exterior or interior? The whole experience of owning is having much less to worry about being “told off by the parents” when you give it back.

Wear and tear is expected. If you get in an accident, you have it repaired just like you would if you owned it. If you beat it to hell then you can expect an adjustment to the value, just as if you tried to sell your own car after treating it that way. As long as it doesn't have collision damage, tires that are bald to unsafe levels, or a trashed interior, the guy at the dealer you hand the keys to is unlikely to care (it's not like the dealer owns it). They're just going to ship it off to auction and the detailers will clean it up and resell it.

Leasing unless the manufacturer is doing something weird to hit quotas is equivalent to financing the car for 3 years and then selling it, so there aren’t big financial advantages one way or another, with financing giving you more flexibility.

Definitely you get more flexibility when financing, which is good if you like to keep cars a long time, or a short time. The latter being a surprisingly common situation. Some people figure a lease is a good way to have new cars pretty often, but don't realize that 3 years is still a fairly long time if you are a car spaz.

It's nice, though, to pre-negotiate the end of the ownership experience, selling cars to private individuals is frequently a great big hassle and trading in is a great way to spend several thousand extra dollars just to avoid that hassle.

Not quite.

Leasing lets me know what the car will cost me up front and I can decide based on those facts.

Buying exposes me to the whims of the resale market when I decide to sell. I also have to deal with the distraction of actually selling the car. If I want to maximize my return, that means private party sales, which means exposing myself to potential fraud and unbounded hassles with potential sellers.

I am generally not a fan of leases but I think they make sense with the changing EV landscape.

I'm right there with you. The rate of improvement is much higher for EVs right now than ICE, and it means they [should] lose relative value much faster (covid issues aside). I solved this by simply identifying the red lines for me. What can I absolutely not live without for any reason? If I can get those and EV improvements, then I'm in. Sadly EVs cannot give me my hard red line requirement yet: range. My cheap Honda Civic gets me 700km+ on the motorway on a single tank, and it doesn't have optimal charging ranges, nor does it lose a significant amount of range in the cold.

I'm waiting for at least 1,000km WLTP range in something like a Model Y before migrating. At current pace this looks to be some time in the next five years.

Out of curiosity, what’s your use case for such a high range? 1,000km is 8-10 hours of uninterrupted driving.

You can already go (almost) everywhere in the US with an EV, and the US is more spread out than, say, Europe. Is this range requirement for the Australian outback or something?

If you live in Texas that level of range is a pretty hard requirement. Going outside of Houston or the 35 arterial of DFW/Austin/San Antonio still requires that level of range. The rural fast charging infra is coming along, but still not where it needs to be to comfortably drive an EV to a lot of Texas without some careful planning.

I live in Denmark where it's prohibitively expensive to own two cars. This is common for most of Europe. So our one car needs to handle not just daily commuting (where an EVs excel), but also the longer trips we regularly take to other countries. These are easily 1,000+km and an EV would add 2-3+ hours to the already long trip.

EVs make a lot more sense when a family can own 2-3 cars and they can rotate their commute car with their road trip car.

The same reason you get a smartphone or laptop. There's always a better one coming, but you need it.

Really, articles hype up changes that tend to be marginal. I'm not saying that Tesla isn't doing good work or that companies don't improve products. But most of the time, there aren't really big improvements. Cars from a decade ago are fine. Sure, I wish I had CarPlay, but it's fine and there's always something new coming.

At some point in your life, you either learn to live with being slightly outdated or you spend a lot of money.

Frankly, some of it is perception. You've talked about the 2021 Hyundai Tucson and I'm not sure if you mean the 2022 model-year Tucson. The 2021 model didn't have a facelift and the 2022 model is actually all new - more all-new than the Model Y update. It's half-a-foot longer, has an additional 6 cubic feet of passenger space, 8 more cubic feet of cargo space, 3.5 more inches of rear-seat legroom. It has a new engine offering 23 more horsepower (14% more). Hybrid and plug-in-hybrid versions are available. Now, you might not care about any of those changes, but it definitely wasn't a facelift.

What does the new Model Y offer? Potentially 16% increased range, but Tesla might just convert that into cost savings for them. Right now, Tesla is reliant on government subsidies for its profits and with those potentially dwindling over time, they might be looking to cut costs more than offer better range. Will the chassis be better? Maybe, but it's also possible that the first few quarters (or years) could have quality issues. Cutting edge stuff can end up cutting you.

Stuff sounds impressive and sometimes hits the market as "meh". I mean, Mazda's Skyactiv was supposed to be revolutionary. Mazda hasn't been making bad cars, but it's definitely "meh". Fuel economy numbers are worse than most competitors now and power isn't amazing either. A Mazda 3 ends up way behind a Honda Civic for fuel economy and on-par for power. All that cleaner-burning, high-compressions stuff kinda leading to nothing. Mazda says that their Skyactiv-Body is 8% lighter and 30% more rigid, but it doesn't seem to be anything to write home about.

If Tesla introduces a new Model Y with those new batteries, it seems likely that the new models will have a similar range, with some lowered costs for Tesla. Maybe 5% better range and 10% cost savings for Tesla (I'd love to be proven wrong on this). Likewise, the chassis is probably going to be a marginal improvement, but the big thing is likely to be that Tesla has never figured out quality control on their assembly and so they're trying to get rid of that - and pass it off as an amazing feature. And I'm not trying to say that I don't like progress and that marginal progress isn't important. Marginal progress adds up over time. I'm just saying that the new Model Y probably won't be as much of an upgrade as it feels right now. I showed that the Tucson had a lot of new stuff, but to you it's just a facelift. If the new Model Y comes out looking exactly the same with the same range, but there might be some chassis difference you can't see (and realistically probably can't feel), are you going to think that's a major upgrade because Tesla altered its manufacturing in a way that doesn't impact your experience of the product? No, that wouldn't feel like an upgrade at all.

Again, I like that Tesla is pushing things forward and that's important, but car companies are always coming out with "all new" stuff that often doesn't feel like a big upgrade when you actually get your hands on the product. The number of times that I've heard about chassis getting more rigid is just incredible. And, to be fair, cars are much better, but if you get a new car every decade or so and ignore the hype train, you're fine. If you get a new smartphone every 2 or 3 years, you're fine. But they're always looking to talk up how this year's version is revolutionary! Those things rarely happen. If you wanted a Tesla, the Model 3/Y was revolutionary insofar as they had an affordable price point. It seems unlikely that we're going to see a lot more revolutionary stuff. That's not to say Teslas are bad. It's just that "3% more range" or "maybe better cornering if you're going 120MPH...except that's illegal" aren't really revolutionary. If there's another Tesla revolution to happen, it's going to be a $20,000-25,000 mass-market car (that isn't nearly as nice as the Model 3). But look at the Model S. Has there been revolutions there?

To me, this article sounds like the type of thing I read about Apple sourcing displays from new companies. It's more about business sense for Apple than it really is about the product in my hand. The new chassis seems like an attempt to overcome Tesla's quality problems (which also seem to get more complaints on the Model Y). The new battery packs seem like a modest improvement over the old ones (Tesla is claiming 16%) which might show up in improved range, but might end up more in cost savings.

There's always something new. Most of the time it's a marginal improvement. Even big changes can feel like just a facelift if you don't get hyped for them.

The tech will slow down eventually for sure - but just like with phones, the difference between a 2016 and a 2021 phone feels somewhat marginal (I’m typing this on an iPhone 7 running the latest iOS and it's quite snappy and responsive) vs doing the same on a 2011 to 2016 phone the difference would be enormous. Electric cars still feel like they are in the “2011 to 2016” stage.

> As someone who just recently got a model Y, seeing reports like this just frustrate me.

I'm not sure if you'd actually notice the difference, and it's Tesla, so you'd want to give them a few years to work out the kinks.

> but when is it a good time to buy an EV? As someone who just recently got a model Y, seeing reports like this just frustrate me

I was listening to Chris Harris a little while ago. He made a great point that over the next few years the cars will go through a similar transition loop as a cell phone went in last 10 years. Every six months your best of breed car will feel old.

I’m dreading going electric, it just doesn’t amuse me at all but if I have to, not before 2025 for sure.

I still don't quite understand how this ends up lighter...

Castings always seem to be at least 5mm thick, whereas stampings rarely are over 0.5mm.

It's the chassis, not body panels.

Thats used for the India's Tata Nano

Is this what they call giga press ? Seem them in a couple of YouTube videos. Or is it something new ?

Why are they not getting serious Osborne effects? I wouldn't want an older model, I'd want a gigacast with the fancy new batteries.

Because they currently can’t produce enough older models to satisfy a demand for them. Try ordering MYLR and see the delivery date

They can sell every car they make. Electric cars in general have pretty high delivery times, Tesla is better there though. Also, it is doubtful whether the initial cars from Texas and Grünheide will have the new battery, likely they start with 2170 cells.

Tesla constantly rolls out enhancements to their car lines, so there is rarely the best moment to buy, other than when you need a new car.

>Elon Musk has called the new Model Y a revolution in automotive body engineering, turning a total of ~ 140 parts into 2 metal parts.

This along with the Structural Battery may seems like a major innovation and provides decent cost savings. The problem is that they are making it like iPhone, the car is non-repairable or extremely costly. I wonder if anyone has a decent argument against this narrative?

As the top comments currently note, the panels are still stamped metal so body work will still need to be done in the event of a non-major collision. The difference is that most body parts they're consolidating are already parts you would never repair anyways, having to replace one would likely mean way more of a cost in man hours and parts than the car's worth, ie. a write-off regardless of this new manufacturing process or not.

Is it just me or does Medium have really poor performance for a site that only delivers text and one image in this case?

Edit: This article scores a lighthouse score of 23 out of 100, wow [1]

And get this: Loading the article through Scribe, an alternative unofficial Medium frontend [3], scores 97 [2]

[1] https://pagespeed.web.dev/report?url=https%3A%2F%2Fscribe.ri...

[2] https://pagespeed.web.dev/report?url=https%3A%2F%2Fscribe.ri...

[3] https://scribe.rip / https://github.com/digitalblossom/alternative-frontends

If you are searching for a Medium alternative to host your blog check out Notaku [0], it can create a blog sites (also docs sites) from Notion content

[0] https://notaku.website

Probably worth mentioning you are the creator of this site?

Why is this the top comment, this has nothing to do with what the article is about?

Because it’s important to remind people that medium is not good.

I wish I could find a way to agree with you that doesn't involve profanity. But here it is:

Fuck medium. I absolutely despise it. Every time I Google for data science techniques I'm inundated with god-awful medium articles and sometimes good medium articles but the paywall is just so stupid. Typically I end up finding the exact same post from the exact same author duplicated on another site.

It represents to me a huge degradation of the spirit of the world wide web as envisioned by Tim Berners-Lee. And yes there are many other sites that also represent this degradation such as Facebook and yes I hate them as well.

Just block their JS. The content survives.

Does anyone have an idea why Medium articles are ranked so well on Google?

The article is blogspam, so might as well complain about Medium.

Why is a good question. But its usually complex enough to discourage chasing answer which is anywhere near the truth.

The ultimate answer is the Universe, simplified version takes into account only brains of people who upvoted which is also unapproachable.

So we settle for stories. Ones that are believable. Say this comment allows us to participate in the discussion without reading the article, because it has nothing to do with it. You of course open comments first, and there it is, comment about Medium (hey I know something about Medium!), which doesn't touch article - great because I have no idea what's in it.

But most whys have uninterestingly many answers. We would love to just get the "main" reason. But how do you differentiate them? Say I commented because I felt like typing but also because I have Internet access. How do you assign weights to those?

Another story reason - it's tech. We do tech. You can't escape the fact that once you have sophisticated patterns associated with something you perceive them. Ornithologist walk through the park is not the same as mine. And psychologist's social interactions are.. different.

Also, why why? How many top comments are directly related to the article and not general topic in vicinity of it? Is it enough of an outlier to be interesting or is your model wrong?

Why did I type all this.

Indignation and fault-finding routinely get upvoted. It seems rooted in human nature (or at least factors beyond our control), so the only thing that seems to work is for moderation to act as a countervailing mechanism, e.g. by downweighting such subthreads. I've done that to the GP now.

It's also why we added this site guideline: "Please don't complain about tangential annoyances—things like article or website formats, name collisions, or back-button breakage. They're too common to be interesting."

https://news.ycombinator.com/newsguidelines.html

I get, on Android mobile in chrome : First Contentful Paint (FCP) 2.1 seconds, it's ok scores 99 points.

For the medium url I get

First Contentful Paint 3.1 s Time to Interactive 18.5 s Speed Index 9.7 s Total Blocking Time 6,160 ms

Horrible.

Reduce unused JavaScript and defer loading scripts until they are required to decrease bytes consumed by network activity.

Even main.js is guilty.

1.9s there.

Serve images in next-gen formats 1.8 s

And very interesting, why does the following happen.

Avoid multiple page redirects 1.56 s

Reduce the impact of third-party code Third-party code blocked the main thread for 5,660 ms Minimize main-thread work 15.7 s Reduce JavaScript execution time 12.3 s

Wow

"Please don't complain about tangential annoyances—things like article or website formats, name collisions, or back-button breakage. They're too common to be interesting."

https://news.ycombinator.com/newsguidelines.html

I think it's very interesting that a proxied version of the site is 4 times more performant than the original

Even so, that doesn't really belong at the top of an unrelated thread.

(I posted a similar explanation earlier today if anyone cares: https://news.ycombinator.com/item?id=29857585)

>requiring custom metal alloys produced in part with SpaceX.

Folks don't realize all his companies (Tesla/SpaceX/Boring Co./Starlink/Solar Energy) are one big company. Yeah they have different names but they operate as one, taking advantage of every possible synergy between them. It is a game changer, everything from material science, communication, energy storage, transportation, etc.

It's like the Borg....a lot of companies and industry won't know WTF hit them.

Reply on news.ycombinator.com