Hacker News new | past | comments | ask | show | jobs | submit login
Ferrari is bricked during upgrade due to no mobile reception while underground (reddit.com)
526 points by a5withtrrs on Oct 12, 2020 | hide | past | favorite | 325 comments

A lot of readers here, and on Reddit seem to be misunderstanding the situation.

The way I understand it, this car was taken in by a garage for 'seat installation'. I presume this means after market seats.

The car disabled itself during installation. The OP assumed due to an anti-temper device, but it could have been any failure mode. Most modern cars with airbags will have sensors in the seats to disable some of the airbags if there is no passenger sitting there. So it is no surprise that any type of work on the seats puts this system in an error state. Especially on high-performance cars it is not unusual that the car immobilises itself when the safety systems are in error state.

A certified dealership will have the computer interface and software to reset the fault mode. For those that do not have this software, Ferrari can remote into the car and reset it for you. This requires connectivity though. This is actually quite a common thing on luxury cars.

This has nothing to do with 'smart' modes, remote hacking capability or disabling a car while in motion. The fact that the manufacturer can remote into their cars, does not automatically mean that the car is vulnerable to remote hacking. This is, in fact, a safety system that kicked in.

The solution here was probably simple: bring the car outside, and let Ferrari do it's thing. Or trailer it to a certified dealership to have it reset there.

> The way I understand it, this car was taken in by a garage for 'seat installation'. I presume this means after market seats.

It was even more basic than that. They were installing a child's carseat for the owner, and I believe it was an anti-theft system that was triggered.

The story reminded me a lot of the rental car that bricked itself because it lost cellular service - https://arstechnica.com/cars/2020/02/driver-stranded-after-c...

Who puts a child’s carseat in a Ferrari?

Someone who has both a child, and a Ferrari.

"This baby can go zero to sixty in three seconds flat."

For anyone wondering, assuming linear acceleration that's roughly 1 g. In a rear-facing seat, that's 1 g eyeballs-out, which should still be perfectly safe for the child, but possibly not for the cleanliness of the seat.

...and if you're wondering what 1g eyeballs out feels like, you can always try looking at your feet.

Shouldn't it be actually ~1.4142g at a 45 degree angle?

You're proposing that gravity is stronger in a direction other than the one straight down to the center?

I think they’re referring to the fact that in the Ferrari, you’d experience both 1g eyeballs out plus the 1g of normal gravity, for a total of sqrt(2) at 45 degrees, when compared with looking at your feet producing 1g down.

Also, basically any boring passenger car can brake at or slightly above 1 g.

Maybe if the acceleration is smooth enough, but the sudden appearance of a new g yanking your head to your chest might be dangerous for a baby.

A performance Tesla can also accelerate to 60 under 3s, and with proper warning to hold their heads, my young kids love the rush.

That's a great way to have to change a diaper. Or your pants.

I hate to sound inflammatory, but just to get everyone on the same page, and because I was wondering the same thing:

I thought Ferraris were show-off cars for the rich, and if you were rich enough to have one, and had a child, then you would (sensibly) own a separate, functional car when you actually intended to drive that child around (or employ some car/chauffeur service or whatever).

Edit: I googled Ferraris with back seats, and got this article which opens by acknowledging that people don't expect them to exist or for Ferraris to be kid friendly, so I think I (and the heavily downvoted GP) can be forgiven for harboring this misconception.

>>Adam Merlin, President at Merlin Auto Group, confidently answers, “Every time I speak about Ferraris being kid friendly, people think I’m nuts. And for the most part, I am. After all, even Ferraris that do have back seats often require the rear passenger to sacrifice leg room, so you can imagine peoples’ reactions when I allude to fitting car seats back there. However, after trying a plethora of Ferrari models, I can confidently say there are models you can choose from, if you have a family and still want to spoil yourself by driving a Ferrari.”


I agree with you, and honestly I should have been more clear here.

I also think there are different definitions of "kid friendly".

Using a Ferrari to take your 3-month-old toddler to a doctors appointment? That seems like a waste.

Taking your 7 year old kid on a drive in the "fun car"? If I had a Ferrari, and a kid, this would 100% be happening.

But you just need booster seats for 7 y/os, which simply lie on top of the OEM seat without risking this kind of incident, right?

I have 0 experience installing a booster seat in a Ferrari, but frankly I would expect any form of child seat to be installable without risking this kind of incident.

I don't have the knowledge to get into any specifics about this. The point is just that there are plenty of valid use cases for having a child seat in a Ferrari.

On that point, I completely agree.

does it have a back seat?

First, some models do. I don't know about that particular one, because (shockingly) I don't own a Ferrari.

But second, that doesn't matter. Most places do not have laws that prohibit children in the front seat of cars, only that they use appropriate child seats. Further, child seats are typically required for _children_, not just infants. IIRC, for example, in Texas, child seats are required for any child under the age of 8, unless they are taller than 4'9". It is totally reasonable for a 7 year old to want to go for a ride in their parents "fancy" car. It's actually very responsible of this person to get a child seat properly installed in order to support that.

The Ferrari in question is a Portofino, which does have back seats, cramped though they are.

except that child seats are mostly safety theater, of course

Absolutely not the case. A 5 point harness is best. A car seat converts two or three point harness to a 5 point. Which is great, because those soft little bodies need as many touch points as is possible to spread the load of force. We know for a fact that the 5 point harness is the best, that is why race car drivers use 5 point harnesses instead of lap belts.

The three point harness is a good compromise between the uselessness of a lap belt and the incovinience of a 5 point harness. But its certainly not as good as a 5. Point harness.

I think the other reason that street cars don't use five-point harnesses (and why using them is illegal in most states) is that they immobilize your shoulders and neck too well, which makes rollovers more dangerous without a roll cage. You can't lean forward when the roof collapses, so you can be paralyzed.

And it's hard to look towards the rear quarters.

Ah, that's another good point. Turning your head happens a lot more in street driving than racing.

And also they're inconvenient and don't auto adjust.

3-point belts are made to be as quick and fool proof as possible so that people actually use them, and use them correctly. Importantly, they auto-retract so that they are worn in the proper position.

When you put on 5-point belts, you have to 4 times as many latches to operate, and you have to adjust all 5 belts -- and in the proper order. If you tighten the belts in the wrong order, they won't protect you properly. It is not intuitive -- I have a car with 5-points and the passengers who manage to figure out how to latch them almost universally tighten the belts wrong.

Yeah, a 5 point would suck if you had a dress on.

Dumb question: how do race cars and their drivers handle "blind spots"?

Edit to answer my own question: apparently they use spotters that radio in to them whether or not it's clear, and (at least in NASCAR) the cars don't have side mirrors but have an extra long rear view mirror.

They also only drive forwards which helps a lot.

When I was a kid on the farm, all the transmissions in the grain trucks were modified so you could only go forward. Too many accidents driving backwards with bad sight lines. Its a strange thing at first but you get used to always parking with a forward exit.

And less of a modification on an older farm -- backing up horses was hard enough that many older farms will already be designed for one-way traffic, with all eg barns allowing you to pull through.

huh, never thought of that before but that makes complete sense. All the farms I ever worked on were newer mostly created out of irrigation projects from the new deal. All mechanical at that point.

And you're not driving 200 mph.

Lap belts are far from useless; they get you probably 80% of the value of a three point harness. Yes, your torso will flop forward, but particularly for a child's body size there's very little for them to hit.

The heads of very small children are quite heavy relative to the child's size. The concern with a baby is not that they'll hit something, it's that the weight of their head relative to their body will result in spinal injuries.

Consider for instance the $34 million verdict in Texas against Dorel, which failed to warn against placing children under the age of two in forward-facing seats. The kid Cayden was properly restrained in the back seat, but the weight of his head alone was enough to result in spinal injuries that have left him partially paralyzed for life.

We used to laugh at my kid when kid tried to reach for things on the floor and then fell on their head due to the weight of the head relative to the rest of the body -- but that's why rear-facing and five-point is important for children under 2,3 years old.

Oh yeah, a rear-facing infant seat is absolutely a difference-maker - from the talk about "child seats" I thought we were talking about forward-facing seats for children above 3, the benefits of which are pretty marginal AFAIK.

If you've ever tried to buckle a toddler into a standard 3 point belt made for adults, you'll see that the lap belt goes directly across their throat.

Except they're totally not, and saying that is dangerous.

Go look at crash tests of "child car seats" that are nothing more than a box ticking exercise which uses the adult seat, versus a properly engineering seat and you'll see what you are completely wrong.

That Freakonomics article is from 2007 and was only pointing out that manufacturers don't test all scenarios - not anything resembling your claim that they're just "safety theater". CDC strongly disagrees based on actual accident data (click risk reduction at the following link):


Wrong. There's about a 70-80% risk reduction for injury: https://pubmed.ncbi.nlm.nih.gov/15094411/

Is there any evidence for that? I know that for example women are more likely to be injured in car crashes because most car seats/interiors are designed around the average man so if be surprised if a human half didn't need to be raised or put in a safer position.

Don’t men get in more accidents?

Yes, because seatbelts are so effective for us, it turns out there is no selective pressure avoid car crashes.

I didn't know that about seatbelts. All these years of careful driving have been wasted on me.

It only looks that way because we get in more on-purposes.

So your "evidence" is a single anecdotal account from over a decade ago?

Actual accident data doesn't back you up...

Some do.


But what difference does that make? Why couldn't you put a car seat in the front passenger seat of a two-seater?

Airbags are not designed with the bodies of children or infants in rear-facing carseats in mind. They're actually really dangerous for them - potentially lethal.

And most vehicles have methods to disable airbags when the item/person occupying the seat can not safely handle it.

But if you want to get real sad, IIRC many airbag systems are not designed with an occupant outside the parameters of a typical adult male in mind. Some modern systems are getting better about this, but if you're too short, too tall, a woman, or other-wise-outside-of-the-norm, good luck.

It's worth noting that here in Australia cars have no such method, and you are not allowed to put a car seat in the front seat.

Equally when travelling as best I can tell all European cars have a method to do this and you can install in the front seat.

So this probably varies by country, but I am not sure which countries have what or where the USA is.

Without specifically checking I imagine that the Australian Design Rules basically wouldn't allow them to add the option even if they wanted to - if I had to guess because you're not allowed to have such a disable button (which you could argue could also be dangerous in some cases if pressed or used when it shouldn't be).

What goes wrong with a woman that is the expected size?

Very few women are of the expected size (5' 9" tall) aimed to represent the average male; most women automatically falls into the "too short / other-wise-outside-of-the-norm" as far as car safety testing was concerned until recently - AFAIK cars made since 2012 should be also tested on smaller, female-sized dummies which also created an incentive to alter how they are built.

Well if that's it then making a list of "too short, too tall, a woman" is double-counting.

Also if you take it as exactly 5'9" then airbags aren't suitable for men either. What you need to do is examine the range it's designed for. It's not the average height that matters, it's what percent of men/women fall inside that range. There might be a significant difference in that percent, or there might not be.

There is no such thing as an "average" person: https://www.thestar.com/news/insight/2016/01/16/when-us-air-...

That analysis has been done, and there is a significant difference in that amount. You can back-of-the-envelope it yourself if you know that the average height of a woman in the US is 5'4" with a standard deviation of about 3.5".

> Airbags are not designed with the bodies of children or infants in rear-facing carseats in mind.

According to FMVSS 208, they're tested using an unbelted 50th percentile size male dummy. That's the reason they deploy with more force and are triggered at lower impact speed had they been designed with belted occupants of varying sizes in mind.

They are tested without belts to ensure the airbag doesn't kill idiots not wearing their belts. But on all modern cars they're designed to work together with the seatbelts.

If you unplug the cable at the bottom of a BMW seat (so it doesn't know the seat occupancy and belts) it will disable the airbag and show an airbag light.

I suspect they were in the process of disabling the airbag when this all kicked off.

I think it’s illegal in some states, but I could be wrong.

I find it unlikely that, if you don't have a rear seat, you cannot legally carry a child.

I do not see any where that's the case.


LMGTFY: https://www.cdc.gov/injury/features/child-passenger-safety/i....

> Buckle all children aged 12 and under in the back seat.

> - Airbags can kill young children riding in the front seat. Never place a rear-facing car seat in front of an airbag.

Many modern cars have the ability to disable the front passenger airbag, for when you do need/want to put the seat in the front.

And, for some cars, the front seat is safer for everyone, including children, thanks to the focus on improving the safety ratings for front seat passengers (which are often advertised) [1]. For example, my car last car had side curtain airbags only in the front.

On average, especially with older cars, I assume the CDC is probably right. Their recommendations necessarily have to be the most useful for the average/masses.

1. https://www.nbcnews.com/business/consumer/study-shows-back-s...

Not all child seats are rear-facing child seats.

No but for tiny little newborns, all are. Something about not having proper muscle structure around spine to handle thrust forward during frontal (most common) crash. Also reason why you should never shake babies or expose them to any form of higher G.

We're changing the child seat to front facing right now - our baby is 10 months old. It can be done little bit earlier but not too much.

You should actually hold off for longer

> In 2018, the American Academy of Pediatrics (AAP) released new recommendations for car seat safety. As part of these recommendations, they removed their previous age-based recommendation that children remain rear facing in car seats until the age of 2.

> The AAP now suggests that children remain rear facing until they reach their rear-facing car seat’s weight/height limits which, for most children, will leave them rear-facing beyond the previous age recommendation.

Source: https://www.healthline.com/health/childrens-health/forward-f...

My wife and I got into an argument about this when it was time to switch our kids car seat so I pulled up their recommendations and read their sources. Their cited evidence doesn't support their recommendation at all (nor did it when they first made the rec in 2011).

Their 2011 rear facing recommendation is based on a retracted article and their 2016 rec is based on the same article's data except using a smaller dataset which they acknowledge is not statistically significant. The only other data cited looks at rear facing Swedish children compared to _unbelted_ Swedish children.

I went back and reread it to make sure I was remembering right. I'm dumbfounded that this was allowed not only to stand as a rec, but was increased (deepened? made stronger?) when they had even less data than they did before. I get that it feels like it should be safer, but that's not science and I would have expected a group making medical and safety recs to maintain a higher standard.


I didn't say they were all rear-facing. lisper asked "Why couldn't you put a car seat in the front passenger seat of a two-seater?" I don't know anything about safety of front-facing or booster seats for kids in front seats. But I think my response was a reasonable answer/example of when you wouldn't want to put a car seat in the front passenger seat.

Ferraris having back seats goes back to the early days of the marque: https://www.roadandtrack.com/car-culture/g6602/ferrari-four-... Not all, of course, but many do.

yes, the Portofino M does have back seats, although they are limited somewhat.


Not only does it have back seats, it has ISOFIX points.

I do. Got an older F430. My son 5 years old loves to sit next to me. And me as well. On the F430 and I believe most other models you can disable the airbag as well. Sometimes safer when the seat is positioned reversed.

Apparently enough people to create a market for seats specifically designed for it: https://www.reddit.com/r/Justrolledintotheshop/comments/j9qn...

My first thought: imagine the geometry of trying to get a child into a car seat through a Ferrari door opening …

It's a convertible so it's probably pretty easy.

A baby, that would be weird. But nowadays you have to put a child in a seat until they are like 10...

You aren’t required to always drive fast in a fast/quick car.

Someone trying to show off their money by driving their Ferrari to the grocery store.

>The fact that the manufacturer can remote into their cars, does not automatically mean that the car is vulnerable to remote hacking

If your car can connect to the internet it's vulnerable to hacking via the internet. My car, which is incapable of connecting to the internet suffers no such potential for abuse.

Theoretically if the systems are properly decoupled and airgapped than the hacking surface area should be quite low. You might be able to hijack the music coming through the speakers, but you should never be able to control the brakes over the internet.

In real life is that what actually happens? Sadly, no. Car companies seem to make a lot of bone-headed systems engineering design decisions (such as having a computer network path connecting infotainment to critical safety systems)

> Theoretically if the systems are properly decoupled and airgapped than the hacking surface area should be quite low.

Since an over the air update is supposed to improve your 0 to 60 times, pretty much all systems must allow to be updated and thus hardly can be airgapped.

How small the surface is is irrelevant, an internet connected car is more of a risk. I don't want all these features in a car.

Something that is connected to the Internet is, by definition, not airgapped.

It's probably on a VPN or a private APN and not the internet.

Just because the car has external connectivity doesn't mean the car has abusable vulnerabilities. Of course, some vendors are going to be better than others when it comes to hardening, pen-testing, and good operational security.

Even if the only allowable connection is the legitimate manufacturer's, the manufacturer itself could be compromised one day (if not already), allowing malicious commands to be sent on the trusted downlink. Or an "oops I found a bug in the code you just pushed to 100,000 cars" could happen accodentally.

It’s unlikely but possible for an over the airway connection to actually be inherently secure. One example is an isolated system that could monitor say engine temperatures and oil pressure and nothing else.

That sort of system is totally _possible_, but not the way any modern automotive connectivity works.

Modern automotive connectivity is via the CAN bus. It is not a given that the system that's internet connected is even able to talk to the CAN bus.

Most are. GM OnStar systems & Tesla I know for sure are designed in a way that a malicious hack of HQ computers + research into car firmware could lead to a push that would kill potentially millions of people simultaneously. If I really dived into the others, I wouldn't be surprised if ford, VW, toyota and others had similar system design issues.

We are just lucky we live in a world where even 'hackers' don't go that low. I hope future regulation will mandate remote access airgaps. I don't think it's a big price to pay for non-entertainment / GPS software updates to require a physical connection

> We are just lucky we live in a world where even 'hackers' don't go that low.

Unfortunately they do go that low. There have been multiple murders committed by ransomware gangs attacking European and US hospitals in recent months. Example: https://arstechnica.com/information-technology/2020/09/patie...

That's more spray and pray damage then a specific hack on hospitals.

Modern ransomware attacks do recon and exfiltration for weeks or months before pulling the encryption trigger.

The attackers knew it was a hospital network and clearly didn’t care that people would die.

How's it going to unlock the car if it's not talking to the important parts?

Sure someone could make a dedicated secure unlock wire or something, but they don't.

> A certified dealership will have the computer interface and software to reset the fault mode. For those that do not have this software, Ferrari can remote into the car and reset it for you.

Ferrari actually flew a technician in to visit this car, who then said "yep, nothing I can do here." Then they had to push it out and tow it.

> The solution here was probably simple: bring the car outside, and let Ferrari do it's thing.

The solution was not simple.


Agreed, I also really dislike this frequent misuse of the term "bricking". Bricking used to mean getting your device into a permanently unrecoverable state, but these days I see it used so often as just to mean "I can't turn it on right now".

Soft brick vs. hard brick is a common (though equally ill-defined) term.

If the user cannot restore it to normal operation using tools that the average user has available, that is definitely a brick though, and I'd argue that it meets even the definition of a "hard brick".

Bricking has never had a solid definition. The most useful definition is a product is bricked when the owner does not have the skills/time/money required to get the product in a working state. The product is a brick to them.

If you flash bad firmware on a device and overwrite the bootloader used for flashing, many would consider this a bricked device since they don't know how to use an SPI flasher to restore a working firmware. For others simply having a device that isnt fixable via a factory reset is bricked because they don't know how to flash via usb.

A car that has become software locked without a way to restore it is bricked imo.

>Bricking has never had a solid definition.

Yeah, the definition is "turn it into a brick"/paperweight, hence the name. A "brick" in unrecoverable, as the name implies.

Unrecoverable is relative. If I wipe out the bootloader on a phone its unrecoverable because I don't know how to flash it directly to the chip and I don't have a copy of the bootloader to even flash. Its a brick because I will never be able to use the device again. If I send it off to the factory in china they could certainly restore the device to working condition again but it would cost more than a new device if they even accept my one off job.

If a DRM system triggers on a car and I can't disable it its now bricked if the dealer won't come and fix it.

So, how is a user to know the difference between “I tried all methods known to me and googled for solutions, but can't turn it on right now”, “the hardware is fine, but nobody can turn it on ever again” and “there’s a hardware problem that means nobody can turn it on again”?

Because if that, I think it’s natural that the concept gets broadened.

Language evolves, it's not set in stone.

Besides, the car might still be bricked, as far as we know? Perhaps the onboard computer is screwed and needs a replacement?

> Language evolves, it's not set in stone.

Yeah, but in this instance I feel like the language is evolving solely because it makes for better clickbait titles. "Bricked" sure sounds scarier than "temporarily unavailable".

Well almost nothing is actually 100% bricked, most phones or motherboards (for bios) can be programmed through some hardware interface. If you brought it to the factory / R&D lab, odds are they would be able to flash most things they designed.

It's just difficult and uneconomical, so isnt offered as a service

You're right of course, but nowadays it's become sort of "some working object has become inert due to software and requires physical intervention"

Exactly. Totally a clickbait headline despite HN comments suggesting evolving language.

This is about as far from bricked as you can get.

> The solution here was probably simple: bring the car outside

They did. Because the car did not have connectivity at the time it detected the issue, it went into full lockdown (judging by the comments, it kept trying for a while, couldn't connect, assumed the worst).

The car performed exactly the way it was programmed to perform. Tampering was detected. No connectivity was found. Lockdown. It will now have to be serviced by Ferrari, a local tech is not enough.

It is difficult to imagine a system that's just as secure but less inconvenient. If you open the door for local 'patches', then you open avenues of attack. Better to do a `rm -rf *` and blow up a firmware chip or two for good measure. Threat vectors are potentially more sophisticated for these cars, if there's a way to locally bypass the lockdown, it will be found.

The target demographic for these cars will not be inconvenienced much. As per the post, the owner went and got his Maserati.

So the takeaway for any Ferrari owners is: make sure there's connectivity if you are trying to perform any changes to the vehicle.

The way I understand it: "smart" everything is getting too clever for its own good.

In your example, why not have a push-button override to allow the car to reset after seat installation?

Soft-everything. I miss my hardware switches.

- signed getting grumpy, getting old man

this is really intended as an anti-theft mechanism, screwing with the car will lock it up and require a call to support to unlock it. So a push-button override would defeat the purpose.

- Undersigned same boat, mid twenties

> The fact that the manufacturer can remote into their cars, does not automatically mean that the car is vulnerable to remote hacking.

I'm pretty sure that that's exactly what it means. What happens when someone steals the manufacturer's authentication keys, or gets on to the manufacturer's VPN?

Security and safety systems in a car are 2 different things.

Being able to reset a car that was immobilised due to a failure in the safety system does not mean that you can steal it, nor does it mean that you could create a dangerous situation. Resetting the system does not mean overriding. If the problem persists, the car will automatically immobilise again are the reset.

> let Ferrari do it's thing. Or trailer it to a certified dealership to have it reset there.

This is a very Apple-like line of thinking. It's not a valid excuse.

Luxury car brands are Apple like. Reputation is everything. They don't want someone modifying the car causing it to fail and becoming a big news event. So many stories about explosive batteries and chargers on phones are the result of some dirt cheap china part being used but the news just reports "Product from corporation bursts in to flames"

The "car seat" is a child safety seat, or booster seat. They were strapping an extra apparatus in on existing hard points, nothing that would bother the airbags at all. It was probably some sort of anti-theft mode.

There was an update to the story. They pushed the car outside but that didn't help so they had to tow it.

> The fact that the manufacturer can remote into their cars, does not automatically mean that the car is vulnerable to remote hacking.

I don’t see how that follows? The fact that someone can remote into your car is exactly what makes it vulnerable to remote hacking.

>The solution here was probably simple: bring the car outside, and let Ferrari do it's thing

Except the self-same failure mode means you can't bring it outside either except via an expensive process.

More generally, I don't know how any of your points should reassure anyone. Put yourself in this driver's place and tell me how dismissive you'd be at such an event.

> So it is no surprise that any type of work on the seats puts this system in an error state.

It's only not surprising if you already know a ton about cars, which is a terrible definition of "surprising" to use here.

I get what you're saying, but:

> The fact that the manufacturer can remote into their cars, does not automatically mean that the car is vulnerable to remote hacking.

I mean, it pretty much does.

"This has nothing to do with 'smart' modes, remote hacking capability or disabling a car while in motion. The fact that the manufacturer can remote into their cars, does not automatically mean that the car is vulnerable to remote hacking. This is, in fact, a safety system that kicked in."

Uh, that should be the other way around: "The fact that the manufacturer cannot remote into their cars automatically means that the car is not vulnerable to [that form of] remote hacking."

> The solution here was probably simple: bring the car outside, and let Ferrari do it's thing. Or trailer it to a certified dealership to have it reset there.

This sounds like it invokes the "right to repair" issue though. Is this "Right to repair, but only if the car's online"?

> "Right to repair, but only if the car's online"

You have the right to repair. But you do need the correct tools to do it. If you don't have those tools, Ferrari may help you remotely, but you'll need connectivity for that obviously.

It looks to me like it's not possible to acquire or make the correct tool, and only Ferrari dealers have it, because it's been restricted by Ferrari. But that's not clear from the information given.

In other words, a classic case of: It's not a bug, it's a feature.

> anti-temper device

I want one.

The updates are pretty hilarious as well. Basically even after sending the tech the car is still bricked at the moment.

Update 1: https://old.reddit.com/r/Justrolledintotheshop/comments/j9ji...

Update 2: https://old.reddit.com/r/Justrolledintotheshop/comments/j9qn...

There was also a story a few years back that is kinda similar in issue. A rental car was taken into a rural area but they couldn't restart the car as there was no cellular signal for the app to re-enable the car. https://twitter.com/kari_paul/status/1229214223227478016

Haha, a few years back? That tweet about the stranded car is from February. I blame you zero percent for perceiving it as such though...

I've had a similar problem with Zipcar rental (car/van rental by the hour using an app) in London multiple times. If you park somewhere without cell signal, you cannot use the RFID card to unlock the door, but if you use the key then the alarm goes off and won't stop. Every time it's happened to me they had to send someone out to reset the security system.

The tech shows up in another Ferrari. God damn, nice company car! Obviously a loaner because the tech flew in from another city, but damn!

Edit: Oops, that's probably the tech in the bricked Ferrari, not in his company car.

This car is not bricked. Please - get a grip on the language being used. The anti-theft device has triggered and the car is in lockdown.

Is there a user control on the device to restore it? If not then it is bricked to that user. Bricked is relative to the users ability to fix it. At some level of skill and money nothing is "bricked" since you could replace or reflash anything.

No, bricked is relative to _anybodies_ ability to fix it without replacing hardware.

Almost all usage for bricked from before it become a mainstream word was meaning that the bootloader on a device had become overwritten or corrupt meaning it is no longer possible to fix using the usual firmware flash method.

This is still fixable if you knew how to access the flash chip directly and had a backup of its content.

Never have I seen bricked strictly mean that hardware has blown up.

LMFAO this pedantic beyond belief. It's bricked, it's a paperweight. They lack the ability to fix it and move it. It's been bricked. Just because there's potential for it to be unbricked doesn't mean it's not been bricked.

Reading through it seems like people are misinterpreting this.

By my reading the car has an anti-theft device that was triggered by the installation of seats. There is a remote system to override the anti-theft device, but that doesn't work because it requires mobile reception.

It's kind of funny, but this doesn't really tell you anything about the dangers of "smart cars" or anything.

It tells us that the people who are designing these things are not considering some of the most basic issues.

I think preventing your car from being stolen is an extremely basic issue - it's a risk that exists at the back of your mind constantly when you drive an expensive vehicle, so yeah, I'm glad the car does this.

The situation in here is more like "haha, ok, we just need to tow this thing out of the garage and try again" - no big deal.

> The situation in here is more like "haha, ok, we just need to tow this thing out of the garage and try again" - no big deal.

No big deal when the reason you can't get a signal is that you're in a parking garage. Rather more of a life-threatening defect when the reason you can't get a signal is that you're driving through an unpopulated desert where there are no cell towers or other people.

Look, I don't disagree with you, it's a problem - but I can guarantee that more Ferrari customers are affected by someone trying to steal their car than there are Ferrari customers driving through an unpopulated desert with no phone signal(and trying to do roadside repair). It's just a question of probabilities, and the first issue seems more important to solve.

Besides, I think there's a certain fundamental misunderstanding here - the car disabled itself because it "thought" it was being stolen. It's no different than triggering a car alarm in any car made in the last 30 years - most systems will kill the ignition until you reset the alarm.

The difference here is that Ferrari can unblock the car remotely for you - but I'm certain it's also possible in-person with some kind of key/button input or some diagnostic device.

It's no different than what 1990s volkswagens used to do, if you removed the battery the radio would be locked until you entered a special pin code, but of course the dealership wouldn't give you the pincode, you had to take it to them and pay them some money. The only difference here is that Ferrari can now do this remotely(if you have signal).

> but I'm certain it's also possible in-person with some kind of key/button input or some diagnostic device.

Seems like you're giving Ferrari too much credit, as the engineer they sent out wasn't able to get it working (eg in-person).

The car needed to be towed (back to Ferrari I guess).

Well, I'm guessing that what's happening isn't "the car doesn't have a mobile connection so Ferrari can't talk to it " - it has literally broken down in some way. Maybe there's even a second side to this story - they severed a wire when installing the seats or something like that.

The engineer was not able to get it working because they allegedly forgot a tool.

Failure modes and effect analysis [1] or let me do a computer security analogy: threat model.

A desert should be considered for a Toyota Landcruiser, for a Ferrari it would be a low probability cell of the matrix. Stuck in a tunnel underground? More likely to happen. In a concrete building with people tampering the car? Even more likely.

[1] https://en.wikipedia.org/wiki/Failure_mode_and_effects_analy...

"Deserted desert" doesn't make much sense for a failure mode for a Ferrari, but off the top of my head I can think of half-a-dozen locations where it would both be super-fun to drive a Ferrari and that don't get cell coverage. All of these less than an hours drive from a major city.

The Silver State Classic is an example of (normally) fairly deserted desert where a Ferrari would be right at home. https://sscc.us/

But then you also have to consider the consequence of the failure. Failure of anti-theft, car is stolen and owner files insurance claim. Failure of vehicle in desolate area, high probability of fatality for driver and occupants.

Ok, but again, that applies to literally any anti-theft system. Like for instance ignition lockout is extremely common on fancier cars, if that system fails for any reason you can't start the car. If you're stuck in a middle of a glacier in your fancy Ferrari then you're screwed. Again, remember that this was triggered when they were taking the seats out of the car,like, come on, that's a "huh, interesting" case , not "huh, someone needs to do something about this" case.

I would wager a guess most of Ferrari clients live in or near a desert. Same for Lambos and other exotics.

> driving through an unpopulated desert where there are no cell towers or other people

These kind of cars can and do fail for many other reasons, basically parts reliability is not a top priority.

Being in a desert with no possible assistance is just a situation you'll want to avoid, anti-theft system or not.

I think you'd have to be kind of nuts (or filming a Top Gear episode, not mutually exclusive) to be driving a Ferrari through an unpopulated desert.

Or just driving the Nullarbor or any other number of very large stretches of road, where someone can experience a high performance car, that tend to be remote and have pretty poor signal.

And before you say that's insane... You'll find plenty of video evidence on Youtube and places like it that this happens. In fact, McLaren sent a car themselves.

And yes, the Nullarbor can be very dangerous if you break down, and there are vast swathes of it with no phone signal, because maintaining anything out there is a losing battle.

Having driven it a couple of times in the last few years I can confirm that the sections without signal are actually very limited. For the vast majority of the crossing you have reception (as long as you are with Telstra).

Maybe, but what do you say when they put the same anti-theft system in a Land Rover, or just the regular cars that the people who live or work in sparsely populated deserts drive?

>> people who live in sparsely populated deserts drive?

Well, by the very definition, there can't be that many people there - so unfortunately they will have to buy something else, or discuss it with their dealership before they purchase the car. When I got my current car(2020 Volvo XC60) the dealership did warn me that if I don't have signal many features in the vehicle won't be available(no idea if it would "brick" itself if I tried repairing it in the middle of a desert, but I'm not interested in finding out).

If one can afford a Ferrari, one can afford a satellite phone/communicator. If one can afford a Ferrari, one also knows that it is far more likely the thing will break the old-fashioned way rather than lock you out via software. Hell, I don't care who made it, cars still get flat tires.

A flat tire doesn't stop you from driving if the situation is urgent enough.

> Rather more of a life-threatening defect when the reason you can't get a signal is that you're driving through an unpopulated desert where there are no cell towers or other people.

Which wouldn't happen. This is an anti-theft system. So are you saying that the anti-theft system would engage while the car is being driven for a while by the correct owner ? If that were the case, that would be an egregious defect that should be corrected.

The car is not disabling itself just because there's no signal, that would be terrible. It disabled itself because it thought it was being tampered with. The fact that not even a Ferrari tech was able to get it going is actually a major endorsement for the system.

> So are you saying that the anti-theft system would engage while the car is being driven for a while by the correct owner ?

If it can happen during a car seat installation then it can happen six hours after a car seat installation when the plug comes loose, or makes proper contact for the first time since the seat was installed, or when you hear something rattle and pull over to try to fix it.

And even if it won't shut down while the vehicle is already in motion, if it does it the first time you stop to stretch your legs, you're still stranded.

> Rather more of a life-threatening defect when the reason you can't get a signal is that you're driving through an unpopulated desert where there are no cell towers or other people.

What happens if your car breaks down for run-of-the-mill mechanical reasons when you're driving through an unpopulated desert with no cell towers or people?

How likely is this failure mode, compared to mechanical failures?

Somehow I doubt this is a concern many Ferrari owners have.

I agree with gambiting. Yeah, it sucks, but just tow your car outside and turn it on again. These things exist for a reason. If it had really been stolen, and there was the work around of 'oh, there's no signal so we'll let the car start and drive off' - then a thief could permanently jam that signal, or remove the electronic device responsible for receiving that signal.

E.g, it's not perfect but it is secure.

OP updated, towing (which is not easy with an activated e-brake) it outside did not work.

[0] https://old.reddit.com/r/Justrolledintotheshop/comments/j9qn...

> ...just toe your car outside...

Hehe I wouldn't normally pick up on typos but this a great one - "tow" your car outside, unless maybe you're planning to kick it out there :-)

Don't know about the mass of this particular car, but I once started an Integra with a dead battery by "skateboarding" it: pushing with the driver's door open and then scooting in to throw it into third once I had enough momentum.

I once came across a couple of people trying to do that with their car on my way home from the pub. Problem was neither of them had any clue how to do it properly. I jumped into the drivers seat and had it running a few moments later.

Only a little while later did it occur to me that what I did was probably illegal in at least 3 different ways.

If it makes you feel any better, a strange woman talked me into climbing into her van at 2AM through an open window to open the door because she had locked the keys inside.

In my defense, I did ask the gas station attendant if he thought it was her car...

Thanks, fixed!

Just tow your car outside sounds like a recipe for losing half a day trying to organise that.

Additionally, even average cars these days tend to have removable towing eyes with easily damaged plastic blanking plugs. The chance of damaging an expensive car doing this seems very high. Especially trying to drag it out of an underground car park. If I owned a towing company I'd take a hard pass on that job.

Exotics are towed all the time. Many exotic rental companies prefer to tow the car for every delivery and pickup. Many owners tow their cars to shows and rallies. (This all seems stupid to me, but I don't own an exotic, so maybe they know something I don't.)

You're talking about towing on a trailer, which is indeed pretty common. Towing by the eyelet is not.

That's how they're winched onto the flatbed though anytime they're broken. The manufacturer provides that eyelet to prevent damage to suspension parts from using straps on a "bad" spot.

I guess if you've got a Ferarri it's a problem for your PA.

Towing company probably has a truck with wheel dollies. I think most of them have them these days.

Like I said in another comment - this just seems like a standard auto alarm trigger, most alarms will disable the car until they are reset. The only thing that makes it stand out here is that normally thanks to the mobile link Ferrari can reset it remotely. If it didn't have that the car would be equally "bricked" and it wouldn't be a story at all.

> it's a risk that exists at the back of your mind constantly when you drive an expensive vehicle

Perhaps due to poor risk analysis or value.

Statistically the Honda Civic and Toyota Camry are much more likely to be stolen; stealing an expensive vehicle is problematic in 1) Security Features 2) Conspicuous to drive 3) Conspicuous to fence 4) Insurance spends more money to find it 5) (in some cases) difficulty to drive.

Bayes strikes again. Just because the Civic and Camry are the most stolen vehicles doesn't mean that they are more likely to get stolen than an exotic car.

From the original thread [1], which people should probably read:

> The fact it had no signal at the time of the tamper safeguard being triggered meant that remote recovery wasn't an option, even when we moved it into the open. Idk, an extra layer of theft protection I suppose.

1. https://www.reddit.com/r/Justrolledintotheshop/comments/j9qn...

This makes total sense, honestly. Jamming a signal would be indeed best done at the time of the (attempted) intrusion.

Yeah, I imagine their security team would have a very unique perspective on this. I imagine it was implemented this way as a response to thefts they saw in the past/envisioned.

people who are designing these things are not considering some of the most basic issues

Is replacing the seats in your car basic?

OP clarifies in the thread that this was for a child seat installation. And I think that qualifies as something that's pretty basic.

In a Honda Accord, yes. In a Ferrari, though?

Rich people don't have babies? Heck, now-a-days they're the only ones can truly afford them.

Some might argue it's the Accord owner that has the money left to raise one (or will maintain enough money left).

Rich people do have babies, just not as many. And they can afford a reasonable car to put a carseat in. In addition to the Ferrari.

That's exactly what the owner did- they went and got their Maserati https://www.reddit.com/r/Justrolledintotheshop/comments/j9qn...

If you have a Ferrari that’s probably not your only car, and you probably drive something more child-friendly when you need to carry your baby around.

Exactly what I wanted to say. In addition, it's like adding a trailer coupling to it - sure, you can do it and it has the power, but that's really not what this car is intended to do.

I don't think it was the act of physically removing seats that tripped it though. It was leaving the doors open.


Not necessarily. Let's consider this for a second.

Having some kind of backup would probably be easier than breaking whatever encryption they're using, and something like this would be a "weakest link" scenario, so anything you put would probably decrease in security.

Besides that, don't really want something a technician can just bring to you, because anything like that can be stolen or sold, so you were probably gonna have to tow it anyways.

I have no clue if that's anywhere close to the actual decisions, but it's not super hard to come up with a thought process that leads to the conclusion that the possible inconvenience of being even more stuck than usual when your car is stolen in a place where you don't have cell service is better than the security reduction of having a backup anti-anti-theft mechanism

Would you prefer it if the system was designed to failed open?

e.g you could override the security system with a signal jammer.

I think the design is correct. If somebody is dismantling the car it should shut down.

Makes me wonder what they forgot to consider in the network security department...

Can you write software of medium scale that is 100% bug free?

Edit: My point is - We shouldn't be making smart cars and TVs because there are always going to be bugs that are not forceable and can cause your device to brick. I wouldn't want a smart microwave or coffee maker.


It is kind of funny that people never thought of that. For the 2020 Census there was a big push to have our interviewing app always connected. Thankfully we did a lot of pushing back, people who live in metro areas have no idea how bad cell coverage can be. And when we have to count the entire US you can’t always be connected. Just seems to me like they made that bad assumption.

and one would think even people who live in metros areas would have noticed that you have little or no cell service when you are underground (on trains or vehicles going through tunnels)-- or when you are deep inside some giant old office building or museum -- or in the middle of a nice sized park.

I live in a major US city and there are still areas within city limits where I can't reliably get cell coverage.

> the car has an anti-theft device that was triggered by the installation of seats. There is a remote system to override the anti-theft device, but that doesn't work because it requires mobile reception.

I’ve been following the thread on Reddit and there’s an added piece you’re missing. The mechanic triggered the anti-theft device so the car went into an initial lock down, that can be disabled remotely by Ferrari. The mechanic contacts Ferrari who can’t override the anti-theft because the car is without signal. At this stage the anti-theft mechanism is also attempting to contact Ferrari, probably with location data to help recover the car, still thinking it has been stolen. But due to the fact that it can’t get a signal out, it assumes the worst and enters full lock down mode that requires some sort of fix that can only be performed by Ferrari themselves.

All as it should be.

Reality - if they had just pulled it outside they could have had it fixed in no time.

TIL the first thing car thieves do is replace the seats.

So what happens if someone tries to steal a car in a cark park?

They can joyride around inside the parking garage as long as they like.

> They can joyride around inside the parking garage as long as they like.

No, the car will be disabled due to the anti-theft system. The anti theft system doesn't need connectivity to work.

It does need connectivity to be disabled remotely by Ferrari. However, that is never needed during normal operation. The owner of the Ferrari modified the car (after market seats, apparently), causing the anti-theft function to immobilize the car as a safety measure.

So Ferris Bueller could have never happened.

Also, Ferrari's are useless in the apocalypse. I mean one benefit of the apocalypse is that maybe you could drive a Ferrari. Nope, they disable themselves when the grid goes down.

Most movie Ferraris are ok, because they either never were real like Modena replica in Ferris or Corvette dressed up as Daytona in Miami Vice, or were gutted by production company with all safeties disabled. The last thing you want during a shoot is a non working car when your burning >$250,000 a day.

Or disable/jam the antenna and joyride indefinitely

The bricking shows that all these “smart” cars with remote software control are liable to hacking (and potentially deadly accidents). We should have a choice to turn them off selectively.

The worst part is that car companies don’t understand software and digital security.


This is false.

According to the reddit thread [1], the anti-theft tamper safeguard (no network required) was triggered. Since it was triggered when there was no network connectivity, network connectivity could not be used re-enable the car.

So, no network connectivity was used to disable or enable the car.

1. https://www.reddit.com/r/Justrolledintotheshop/comments/j9qn...

Which part is false?

The remote control wasn't used here but it normally works and they tried to use it. So it does show that these cars could be hacked remotely.

This part:

> The bricking shows that all these “smart” cars with remote software control are liable to hacking

No remote control or wireless communication was used here, so this can't provide evidence that it is liable to hacking.

The bricking shows how deeply integrated the remote control system is, and that it can brick the car.

And since it's remote control, it can be hacked (or Ferrari themselves can be hacked).

Therefore we can conclude that the bricking shows how badly the car can be affected by remote hacking.

It doesn't matter that wireless remote communication didn't happen here. The wireless remote control system was demonstrated. And the comment is making claims about the wireless remote control system. The only way the difference would matter is if you think they're lying about the ability to send signals to the system.

We don't know if it can be enabled remotely from this story. There is no evidence for that beyond your speculation. There is only evidence that it cannot be enabled remotely. This is why I said your claim is false. Looking at other instances, or the documentation, will probably show that it can be enabled remotely, but that evidence is not here.

But, it appears your perspective is that if remote control operation exists, so it can be hacked. This is proof that remote control operation exists, so it can be hacked. Framed in that perspective, sure I guess.

> There is only evidence that it cannot be enabled remotely.

Wait, by "enabled" do you mean making the car stop working, or making the car work again?

Either way, that's wrong.

For making the car work remotely, the fact that they tried and say it would have worked if it had a cell signal is strong evidence that they can in fact do it. Why would they lie about it? It's evidence they can, not evidence they cannot.

For making the car stop working remotely, this situation doesn't give any direct evidence either way. It's not evidence they can, but it's also not evidence they cannot.

And considering how these devices are usually designed, if they can flip the car one way they can probably flip the car the other way. So it very likely can remotely make the car stop working.

> But, it appears your perspective is that if remote control operation exists, so it can be hacked.

Is that wrong? Even if the system itself is magically perfect, surely Ferrari itself can be hacked. I can't imagine someone credibly claiming otherwise. So I take it as a fact that a hacker can send commands to your car. Whether those commands can shut it down... I'll just say that if it even resembles competitor systems then it's a strong yes.

Imagine doing that on a busy public highway for a wired.com article ...

'we did it in a safe way'

They had no control over the cars behind, anything could have happened.

Their attack could have been demonstrated on a parking lot or private road.

Their information did need urgently sharing, and while I understand the common need for a bit for drama, this was utterly unsafe & for the good of everyone I hope they've pulled their heads out of their asses since making that video!

Just for clarity because this Jeep example comes up quite often when talking about cars:

I don't know any vehicle with connectivity (other than Jeep and maybe Ford?), which does not have safety critical CAN/FlexRay buses segregated from driver facing 'infotainment' systems.

What that means is that the network bus in which your 'compromised' infotainment system is able to operate is completely separate from Engine, ABS, AEB, ESP, Airbags etc.

The solutions vary but there is usually a physical gateway that prevents a passthrough MITM attacks, so you cannot simple send a message frame from your infotainment pretending to be an AEB module requesting emergency braking to your ABS system.

> The bricking shows that all these “smart” cars with remote software control are liable to hacking

No, it does not.

The way I read the article, the photo is taken at the installation facility where (presumably aftermarket) seats were being installed. The car disabled itself during installation. It was not driving. It was also not "hacked" remotely. I assume the installer also had the key at hand.

A certified dealership will have the computer interface and software to reset the anti-theft mode. For those that do not have this software, Ferrari can remote into the car and reset it for you. This requires connectivity though.

This is not bricking. And really has nothing to do with smart devices. I assume this device could be recovered through the usual wired diagnostic port means, so the reception issue is an annoying situation but not anything close to “bricking”

The chances of attempted theft of a Ferrari are likely larger than the chances of someone installing a child seat so this looks like it was implemented ok, though the recovery could have been a bit more elegant.

That's not what bricking means. Bricking means -permanent- disablement.

That's an overly pedantic definition of bricking. The car's in state where is is unusable without intervention from the manufacturer. There's no technical difference between this an a 'bricked' motherboard, cell phone, or other piece of consumer electronics: the hardware is not destroyed, but it's unusable unless the manufacturer is willing to fix it.

The only difference between this bricked Ferrari and a phone bricked by a bad firmware update is that the Ferrari is worth enough that it's worth while for the manufacturer to fix it.

Since this is the intended operation of a tamper induced security lock, I think "locked out" is much more accurate than "bricked" (at least according to the reddit thread).

Keep on supporting the clickbait headline writers - for those of us around a bit longer we know what bricked means. And yes, actually bricking can happen, and no, it does not usually involve a quick and easy reset by a DEALER tech.

If you want bricking, check out Samsung's work around the Note7. "prevent all U.S. Galaxy Note 7 devices from charging and will eliminate their ability to work as mobile devices" - that turns these devices into actual bricks with no future utility.

If establishing a connection to the manufacturer is all that's needed to restore the car (which was the case), there's no way that is bricking. That's like saying my Amazon Alexa is bricked when my wifi goes down.

Analog cars uber alles :)

Jokes aside, cars have not been analog anymore for decades ;-(

Most stuff (even stuff that looks analog like the speed gauge) now runs thru buses and OBD allows you to tap into almost every controllable aspect of the car if you know the 'right api' for the device.

As long as the control systems are properly air-gapped you should be fine. (doesn't work for jeep apparently due to cost savings... https://www.wired.com/2015/07/hackers-remotely-kill-jeep-hig...)

won't stop someone physically breaking in and connecting something to the bus but the alarm should sound in that case.

Well, it's a 2-step process. It's hard to find a car without remote-unlock. From there, you have access to OBD.

OBD port is +12, GND, TX and RX pin if you just want to connect to 1 bus. Shouldn't be too hard MITM that if you have physical access.

1. Remote unlock - you need to hack that... good luck...

2. ODB port requires physical access which means either alarm or 1... under pressure since owner may see you and call cops.

Now you can do something else on cheap asshole car manufacturers like Jeep and bluetooth/wifi hack media center then get on ODB via that because cheap idiots said its a good idea to have common bus.

Iunno, #1 followed by #2 has been happening with some frequency with high end vehicles.

But moreso to steal them in the middle of the night and have in a sea container within 24h.

Then again, if it was being done to install a MITM OBD device... it’s a lot less obvious than waking up to your car gone.

> It's hard to find a car without remote-unlock.

Unplug the antenna?

And someone that's going to MITM your OBD could also MITM your ignition wire on an analog car.

There have been cars where the bus connection to the ABS system was accessable in the wheel well.

Yeah, but you need physical access and unlocked car... you can cut the breaks too if you have that...

Totally. The last analog-ish era of cars was the mid 2000s.

Serious question: I think there is a market for a dirt-simple, stripped down car like the original Jeep. If capitalism is so swift, why is the product I want to buy not on offer?

Answer: the market is captured, and controlled via government regulation, by corporations that do not see individual owners wrenching on their cars as advantageous.

Follow-up: how do we restore capitalism to our (nominally) free market?

You can frame it as regulatory capture, but to a large extent, it's more about environmental and consumer protections that the auto industry has spent decades fighting.

As somebody who drives an ancient car, I'm keenly aware of its lack of safety features. The "american steel" argument falls flat; my car was made before the concept of crumple zones, much less airbags -- driving this boat, I'd be seriously injured or die in a head-on collison with a Yaris, where the other driver and passengers would probably walk away. Unless my gas tank explodes and engulfs us all, that is.

If everybody was driving pre-1984 vehicles, the emissions would result in horrible smog, bringing most cities on par with Beijing.

The world must bring carbon emissions down to prevent severe climate change. Maybe lax safety standards could contribute to that; but environmental regulations must not go.

Follow up: the "free" market that you propose is not actually free, but ignorant of externalities that the modern market has limited facilities to account for.

Is throttle and steering by wire a safety feature? No. Its objectively worse than having the physical linkage, and manufacturers turn to it out of ease of design.

Sorry, but I'm not sure what part of my comment that you're responding to.

Maybe you're rebutting my parent? Because I've never heard of a regulation towards drive-by-wire, so I suspect this particular aspect of automotive design is a result of a free market.

While I could see steer-by-wire driven by lowered production costs (easier to share parts between LHD and RHD models), I could also see it indirectly driven by weight savings to meet fuel economy regulations.

No pinion shaft required. No pinion shaft cutout through the firewall means the body can be made with less steel.

Belt-driven pumps are moving to electric-drive for efficiency, but when you're switching from hydraulic assist to motor-assist, making the whole thing 'by-wire' and motor driven reduces complexity.

Sure, I can agree with the weight savings.

Having hit a pothole recently, I've been driving with a bit of a lean... this could be extremely easy to fix in software, and I suspect that automatic recalibration would be pretty simple to implement. As it stands, I'm waiting for nice weather on a weekend in the Fall, and I'll probably end up at the mechanic.

And economy of scale doesn't just apply to LHD/RHD, but the parts necessary to implement drive-by-wire can be shared between makes and models -- a Fiat won't need as much torque to steer as a Peterbilt, but in theory a huge amount of that system could be shared. That can significantly reduce engineering costs as well.

If your car is leaning bad from a pothole the last thing you'd want to do is artificially correct yourself out of thinking you need to get your wheels aligned.

Meanwhile, I'm thinking about how lazy cars are making us and tinkering with disabling the power-steering pump.

But with drive-by-wire, we could simulate all kinds of scenarios in software and switch between them.

Throttle-by-wire helps with emissions and fuel consumption, technically making cars safer for all of us pollution-breathing humans: https://en.wikipedia.org/wiki/Electronic_throttle_control

Not seeing where in the article it states this explicitly. How does it work to help emissions? Compensation for drivers who have a lead foot?

With a traditional, mechanical throttle linkage, there's a momentary spike in emissions caused by a too-rich or too-lean combustion condition when you suddenly lift off or step on the throttle because the engine computer and fuel injection system can't compensate quickly enough. Throttle-by-wire allows the manufacturers to smooth the throttle position changes. Even when you snap your foot off the pedal, the throttle valve is now made to close over some significant fraction of a second, slowly and gradually enough for the fuel mixture control to keep pace.

Well for small quantity production there are exceptions. Like Pagani and Koenigsegg have a ridiculous high Co2/km brand average because of low quantity. So it's possible.

Also miss ABS/TRC/ESC that's done by computer.

> why is the product I want to buy not on offer

Because the folks that ask this type of question are few and far between, and rarely serious. Otherwise, at least one manufacturer would have brown diesel wagons with manual transmissions. ;-)

Maybe 1 in 1000 people wrench on their own car. Manufacturers don't cater to that. They design cars to be appealing to the other 999, and they design the mechanical bits to be as cheap as possible to assemble in a factory.

Maybe 3rd printing can allow some sort of Renaissance.

Alternative answer: because the market for what you want is too small relative to the cost of producing the thing you want.

Except that the cost of producing that is not actually high. Individuals have built cars out of parts in their own garage. You don't think there is anyone willing to make millions of dollars by producing a thousand of them for the thousand people who want them?

But not if the fixed regulatory overhead is equal to the margin on twenty thousand cars.

>Except that the cost of producing that is not actually high. Individuals have built cars out of parts in their own garage.

"Individuals have built cars out of parts in their own garage" doesn't really prove that the costs for producing a car isn't high. They're most certainly enthusiasts, and they could have plowed tens or hundreds of thousands of dollars worth of parts/labor into it.

People are clearly willing to pay tens of thousands of dollars to buy a new car. That's how much they cost.

I think the main thing is safety.

On the face of it I'm with you, but then when it came down to it I'd be thinking well hang on this actually doesn't seem as safe (or at least not as thoroughly tested to determine if it is) as these other mass-produced things, and wouldn't end up buying it.

If it was easy to make something as safe as a $modern_run_of_the_mill_family_car and that looked 'normal' (i.e. curvy, I suppose, not weird slabs of barely bent sheet metal) I think there'd be loads of Kickstarter campaigns 'basic'/'maintainable'/.. cars.

Making the car safe and making the car comply with safety regulations are two different things.

It's possible and even likely that we have rules which are unnecessary or not worth the cost for small production runs, but which apply to them anyway, because companies doing small production runs don't have lobbying clout and companies doing large production runs are amortizing the cost over hundreds of thousands of cars and don't want things to be easier for smaller competitors.

-Because you need volume to make car production economic. We're a small and, hence, neglected minority.

The product you want already exists, though - it is commonly referred to as the 70 series Land Cruiser. :)

The other answer could be lack of scale. These aren’t artisan goods that someone can easily try offering on a whim, these magnificently complicated machines are only affordable because of economies of scale/mass manufacturing. You can probably find someone to custom build what you want, but it’s not going to be anywhere near cost competitive with a standard model. Maybe you can get close by using stock parts from other cars.

Or maybe this is a market opportunity, and you should set up a manufacturing line for stripped down jeep-like vehicles.

Probably won’t be road legal, though, because they’re horrifically unsafe in a collision compared to modern cars.

There are dozens of car companies, so the market is working. Government regulation exists because your desire to buy a “simple” car impacts my ability to breathe. You can buy very simple vehicles for off road use.

Fixing broken cars became a cultural thing because cars really sucked and you had to waste a fair portion of your life fixing them. Things like emission regulation brought metrics and standards to the industry and resulted in better cars. The typical lifespan of a car is 2x what is was in the 80s (200k miles or 12 years).

> your desire to buy a “simple” car impacts my ability to breathe

This kind of hyperbole is exactly why we have the president we do.

When I was growing up in NYC, first task of spring cleaning was washing the soot from cars off the side of the house.

It’s not hyperbole. Clean air is something we have because of regulatory action.

While we're doing this can we please also bring back small-screen smart phones and physical buttons and knobs on appliances?

Or the market is captured by customer hostile brand focused lifestyle companies that make money on after market spares.

Some of this repeats what others have said; modern safety and efficiency requirements mean the car needs more electronics. When they fail, you're SOL, but modern cars are pretty reliable, making up for part of the desire for consumers to fix them.

Crowd source a Build-Your-Own-Car model kit?

It's called wikispeed. Search for it.

My guess is the product you want to buy is on offer, but not at the price you are willing to pay. Do It Yourself is the market solution for people who want something different than what is offered in the mass market. The internet has enabled truly insane levels of DIY capability. Capitalism is not a charity. Therefore people that want products that cannot be sold profitably in the mass market can make it themselves, or buy it from the hobbyist/enthusiast market.

If this product is on offer, then their word-of-mouth marketing isn't getting it done.

The 70 Series Land Cruiser was mentioned above.

Please explain how this is in anyway due to the/a/any government.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact