Automotive Chip Shortage is as much about the "type of chips" as it is about overall capacity

[leadeater]

20 minutes ago, Levent said:

ECU = Engine Control Unit, which strictly controls the engine.

TCU/TCM = Transmission Control Unit/Module, Only on DCT/Auto/CVT vehicles and only controls transmission behaviour.

BCM = Body Control Module, controls user inputs, reads non engine related sensors (door ajar, AC temperature, air quality etc)

Infotainment system on cars are merely a IO devices, everything is done on BCMs. EVs differ in some ways but they are not that different. You can literally run most 1-2-3-4-5-6 cylinder engines with an arduino (Check out Speeduino if you are interested), as you said ECUs are relatively simple devices, they have coil outputs, fuel injector outputs, ignition maps, fuel maps, boost control for boosted vehicles, sensor inputs, throttle input and throttle body output (in case of DBW) and a canbus. Cruise control on every combustion engine car I know of are controlled via BCM, not the ECU or the infotainment, as I said ECU strictly controls engine, nothing else. Information is carried via CAN to the BCMs, then to ECU.

Although not deep and sometimes not always fully correct the tangential knowledge you get from watching Cleetus, KSR (probably the best for this), FasterProms etc is quite good for picking up bits of information around this. Slight issue is things like Holly EFI combines some of the features an abilities in to a single unit, you still need all the sensors though, more advanced actual "road trim" probably does need the expansion CAN module.

[igormp]

4 minutes ago, StDragon said:

It's not just the ECU, but controllers in the transmission, breaking system, sensors, etc. All of which interconnect along the CAN bus. So when talking about the "automotive industry", the real question is which supply chain/s are being effected here?

Also, most cars have multiple CAN buses, K-line for some simpler stuff, and some newer cars even have ethernet (single pair, automotive) for some high-bandwidth stuff.

[Levent]

3 minutes ago, leadeater said:

Although not deep and sometimes not always fully correct the tangential knowledge you get from watching Cleetus, KSR (probably the best for this), FasterProms etc is quite good for picking up bits of information around this. Slight issue is things like Holly EFI combines some of the features an abilities in to a single unit, you still need all the sensors though, more advanced actual "road trim" probably does need the expansion CAN module.

Holly has some units that offer modern replacements for non ODB cars (pre 1990 IIRC, everything on those cars are independent of each other, for example in older cars AC actuator/clutch is wired straight into dashboard), just like you said those usually have have their own expansion modules that run via canbus so that you can add a tcu, extra sensors or simply if you want to run a modern engine in a old car. Great stuff, Haltech also has some nice videos on youtube.

[CarlBar]

1 hour ago, RedRound2 said:

I am aware of the different standards and diffierent minimum requirements for automotive, military or space. But is it mentioned anywhere that any new tech needs to be tested for a minimum of 4 years before it hits the road? That's what I wanted to know. I get the temperature, durability, reliability aspect of the certifications. That I assume you can also get or request to get with 7nm or 10nm processors today from Intel, ARM and Nvidia -  atleast as per this news

 

 

For 4 years of operation was just an example, i'd expect it has to be rated for longer. It's safety critical and it poops out after a couple of years reguarly the regulatory agencies are going to ask the auto manufacturer to show that they did due diligence and tested it in a reasonable manner. If they haven't then said regulatory agencies will rip the offending company a new arsehole.

 

if your home PC fails, or a datacenter gets knocked offline, generally no one's going to die because of it, (And if they do agencies will ask important questions about fail-safe states and redundancy in safety critical infrastructure respectively), it will be hella inconvenient but no one's at risk. But safety critical stuff that could kill people? Oh boy are you in deep shit if you screw that up.

 

As an example most, (probably all but not 100% on this), fly-by-wire systems involve multiple error resistant computers that cross check each others work and reject erroneous outputs on a majority basis precisely because of that. Car electronics don't generally have to be up to the same standard, but they still need to be pretty serious anywhere they're safety critical.

[Dracarris]

22 minutes ago, CarlBar said:

fly-by-wire systems involve multiple error resistant computers that cross check each others work and reject erroneous outputs on a majority basis precisely because of that.

They actually implement and compile the same program on different microcontroller architectures, chips and compilers. And they all have to agree on the results, otherwise the autopilot immediately disconnects.

 

And yet, there still is this issue with the Quantas Airbus where one of the Inertial Reference Units sent mis-labeled data on a bus intermittently. Short enough for the AP to not disconnect but to send the airplane into really, really dangerous flight paths. They could trace this back to a CPU actually producing wrong results/control flow on silicon level but not reproduce it, this whole thing is not really solved as of today.

 

I also asked myself what you do with an all fly-by-wire airplane when all power is lost. Usually it just becomes a very heavy glider. But in this case, sure there are batteries that keep essential instruments and actuators powered for a short period. If you haven't managed to restart an engine or the APU when the batteries are dead, you are all out of luck?

[leadeater]

1 hour ago, Levent said:

Haltech also has some nice videos on youtube

Haltech is much more popular around here in Aus/NZ with our XR and Commodore series of cars, I think also on many JDM cars too which is very popular here. LS isn't actually very popular engine platform here.

 

Mighty Car Mods is another good watch, and local to me although Aus and not NZ

[CarlBar]

1 hour ago, Dracarris said:

They actually implement and compile the same program on different microcontroller architectures, chips and compilers. And they all have to agree on the results, otherwise the autopilot immediately disconnects.

 

And yet, there still is this issue with the Quantas Airbus where one of the Inertial Reference Units sent mis-labeled data on a bus intermittently. Short enough for the AP to not disconnect but to send the airplane into really, really dangerous flight paths. They could trace this back to a CPU actually producing wrong results/control flow on silicon level but not reproduce it, this whole thing is not really solved as of today.

 

I also asked myself what you do with an all fly-by-wire airplane when all power is lost. Usually it just becomes a very heavy glider. But in this case, sure there are batteries that keep essential instruments and actuators powered for a short period. If you haven't managed to restart an engine or the APU when the batteries are dead, you are all out of luck?

 

AFAIK there's a ram air turbine that automatically deploys, relies on the planes forward motion, fairly hard for it to fail but every once in a while circumstances will line up.. Bear in mind this has been an issue since well before fly-by-wire. If the Hydraulics went out you'd had it too generally. Again redundant, again every once in a while events lined up and they all failed.

[Dracarris]

1 hour ago, CarlBar said:

AFAIK there's a ram air turbine that automatically deploys, relies on the planes forward motion, fairly hard for it to fail but every once in a while circumstances will line up

Right, I forgot about the RAT which should deliver enough power to supply essential instruments (I am not sure about hydraulics).

 

On the A380 the RAT apparently can supply all essential electrical systems required for flying and landing.

 

However, many early planes with mechanically controlled control surfaces use hydraulics only for assistance (like power steering in a car). Should hydraulics fail, you can still use (a lot of) muscle force to actuate control surfaces as long as you are not in a flight path where aerodynamic forces are so high that it's not humanly viable (e.g., in the B737 Max erroneously commanded heavy nose down).

[RedRound2]

4 hours ago, Levent said:

ECU = Engine Control Unit, which strictly controls the engine.

TCU/TCM = Transmission Control Unit/Module, Only on DCT/Auto/CVT vehicles and only controls transmission behaviour.

BCM = Body Control Module, controls user inputs, reads non engine related sensors (door ajar, AC temperature, air quality etc)

Infotainment system on cars are merely a IO devices, everything is done on BCMs. EVs differ in some ways but they are not that different. You can literally run most 1-2-3-4-5-6 cylinder engines with an arduino (Check out Speeduino if you are interested), as you said ECUs are relatively simple devices, they have coil outputs, fuel injector outputs, ignition maps, fuel maps, boost control for boosted vehicles, sensor inputs, throttle input and throttle body output (in case of DBW) and a canbus. Cruise control on every combustion engine car I know of are controlled via BCM, not the ECU or the infotainment, as I said ECU strictly controls engine, nothing else. Information is carried via CAN to the BCMs, then to ECU.

 

3 hours ago, StDragon said:

It's not just the ECU, but controllers in the transmission, breaking system, sensors, etc. All of which interconnect along the CAN bus. So when talking about the "automotive industry", the real question is which supply chain/s are being effected here?

 

ECUs at least the part of the world I am from generally refers to the "Electronic Control Unit". Basically an umbrella term everything you all described above

 

 

Specifically all electronic control units are refered to as ECUs generally. If we're talking about engine, it usally will be eplicitly mentioned.

And specific to this topic, we're talking about the tiny processors that does very simple things inside cars. Unlike most of ther softwares we generally talk about, it isn't such a complicated system to test. All you have to really do is an envrionement test

 

3 hours ago, CarlBar said:

For 4 years of operation was just an example, i'd expect it has to be rated for longer. It's safety critical and it poops out after a couple of years reguarly the regulatory agencies are going to ask the auto manufacturer to show that they did due diligence and tested it in a reasonable manner. If they haven't then said regulatory agencies will rip the offending company a new arsehole.

 

if your home PC fails, or a datacenter gets knocked offline, generally no one's going to die because of it, (And if they do agencies will ask important questions about fail-safe states and redundancy in safety critical infrastructure respectively), it will be hella inconvenient but no one's at risk. But safety critical stuff that could kill people? Oh boy are you in deep shit if you screw that up.

 

As an example most, (probably all but not 100% on this), fly-by-wire systems involve multiple error resistant computers that cross check each others work and reject erroneous outputs on a majority basis precisely because of that. Car electronics don't generally have to be up to the same standard, but they still need to be pretty serious anywhere they're safety critical.

Is there any proof for this 4 year or more claim? Coz I know for a fact Tesla doesn't and they're very open about what the cutting edge hardware they are using. And I highly doubt anyone else does either. Unless you're talking about some super complex system (which honestly im not sure what such complicated system is there in a car apart from self driving) testing ECUs aren't such a huge deal that people are making it out to be. 

 

Flights go by a entirely different standard. That and rockets are one probably the most scrutinized technologies. For those cases, I understand. But for cars, nobody does this sort of extensive testing and honestly in this day and age with better manufacturing and testing procedures in place, its probably not even needed.

 

And most cars do have a mechanical fallback system in place if the electrical components do stop wroking. Steering is still mechanical and so are brakes.

[leadeater]

28 minutes ago, RedRound2 said:

testing ECUs aren't such a huge deal that people are making it out to be

It is because of regulatory compliance reasons. It's not hard testing, it's just long winded and lots of paper work and a lot of evidence required.

 

You have to document all the fault conditions, what is supposed to happen under each condition and then evidence that with test data (many sample points). You have to do this for each one and there could be hundreds. And you have to make sure it does this at both temperature extremes as well, and then after so many running hours.

 

Vehicle safety testing is still a huge deal and always will be, that's why platform updates usually take around 5 years of development, because of the giant mounts of regulation and paperwork required and this does extend to the electrical systems of the car not just structural and emissions.

[RedRound2]

24 minutes ago, leadeater said:

It is because of regulatory compliance reasons. It's not hard testing, it's just long winded and lots of paper work and a lot of evidence required.

 

You have to document all the fault conditions, what is supposed to happen under each condition and then evidence that with test data (many sample points). You have to do this for each one and there could be hundreds. And you have to make sure it does this at both temperature extremes as well, and then after so many running hours.

 

Vehicle safety testing is still a huge deal and always will be, that's why platform updates usually take around 5 years of development, because of the giant mounts of regulation and paperwork required and this does extend to the electrical systems of the car not just structural and emissions.

We're talking about multi million dollar corporations here. Not a group of 5-10 people. They have specific units just dedicated to validation and testing. Like I dont even know what their job is if they don't really do any of this

Second, nobody has still managed to answer this - how is Tesla able to rapidly change computers, be close to cutting edge tech, without too much of all these "burdens"

 

Again, I am not talking about some complex system with over 1000 different data points. It's a damn ECU that an Arduino could do. Platform updates involves so much more of ground up redesign and engineering that requires a complete thorough testing.

 

The answer is pretty simple. If it aint broke dont fix. Why do you think infotainment systems in cars just have sucked so bad, even though we've had smartphones and tablets for a long while now? They're just didnt care. And they just didnt bother to upgrade their aging processor node. Most auto manufactures have focused more on car performance and comfort until Tesla came along with their software first approach

[leadeater]

37 minutes ago, RedRound2 said:

Again, I am not talking about some complex system with over 1000 different data points. It's a damn ECU that an Arduino could do. Platform updates involves so much more of ground up redesign and engineering that requires a complete thorough testing.

Yes and that includes the ECU if you update it to a new model not used and gone through the testing. And no an Arduino could not do it, it'll immediately fail environmental test conditions, again complexity of the task aka the code or compute power isn't really a factor it's making sure it actually does it ALL the time, as best as possible.

 

I would never put my life in the hands of a Arduino control car ever.

 

37 minutes ago, RedRound2 said:

We're talking about multi million dollar corporations here. Not a group of 5-10 people. They have specific units just dedicated to validation and testing. Like I dont even know what their job is if they don't really do any of this

That doesn't change how much testing is required or the complexity of testing. Testing something to the safety standards of vehicle regulations I would say is more complex than designing the ECU in the first place, if not more complex at least more time consuming (person FTE equivalent).

 

You're saying it's not a complicated system to test and your reasoning is these are huge companies with many employees and billions of dollars, all facts that have nothing to do with testing complexity.

 

37 minutes ago, RedRound2 said:

Second, nobody has still managed to answer this - how is Tesla able to rapidly change computers, be close to cutting edge tech, without too much of all these "burdens"

Telsa has never rapidly changed computers. HW1 was Tesla designed hardware and software (2014-2016), HW2 was Nvidia designed hardware (PX 2) Tesla software (2016-2017), HW2.5 was Nvidia PX 2 again with Tesla software changes, HW3 Tesla designed hardware and software (2019-).

 

I would hardly call any of this rapid when they have so few models of cars and do not update the car platforms in the same manor as other manufacturers.

 

 

37 minutes ago, RedRound2 said:

The answer is pretty simple. If it aint broke dont fix. Why do you think infotainment systems in cars just have sucked so bad, even though we've had smartphones and tablets for a long while now? They're just didnt care. And they just didnt bother to upgrade their aging processor node. Most auto manufactures have focused more on car performance and comfort until Tesla came along with their software first approach

No the answer is car platform update lifecycles. Traditional manufacturers do not change hardware or software on existing in production cars, Tesla does. So if you are buying a 2019 year model car that is a 2015 platform model then you have 2015 technology in it not 2019. Tesla has and started with even less models of cars and also restricted which markets they sold in to lower the burden of regulation compliance, an actually smart thing to do.

 

But may I remind you I was replying to you about your comment that testing ECU isn't complex, which it is, and there is a ton of regulation compliance required for vehicles which covers this and the actual business as a whole (IATF 16949 as one example). You think it's simple because the function tasks being carried out is simple, that's just an assumption you've made.

 

Vehicle design and safety is literally life and death and somehow you think the testing required for components like the ECU isn't complex an laborious? I work in the public sector, if something can be made complex and nonsensical they'll do that and double it.

[staticpage]

17 hours ago, Kisai said:

For last 12 years most like Honda's and many others are Electric Power-Assisted Rack-and-Pinion Steering (EPS)  how do you think Vehicle Stability Assist (VSA) works? My 2000 civic came without any PS as do many lightweight "sporty" cars, it is a little heaver than normal PS civics but it was designed to be easier to turn at rest. They make a 12v electric one now you can swap in. Most of what PS does is just convenience and adds a nice factor to the car. Like one part of it was making the steering wheel smaller diameter to make crashes safer for legs but a smaller wheel is harder to turn.

 

 

[igormp]

1 hour ago, RedRound2 said:

And specific to this topic, we're talking about the tiny processors that does very simple things inside cars. Unlike most of ther softwares we generally talk about, it isn't such a complicated system to test. All you have to really do is an envrionement test

Those tiny processors nowadays are pretty powerful by themselves, since we already have a wide amount of dual core microcontrollers out there with caches and whatnot.

Along with that, you also need integration tests since even an entry level car has over 40 ECUs and more than 5 different communication buses. High end ones easily have over 100 ECUs and 10 buses.

 

This is a little outdated, but gives a rough idea how large a car codebase is: https://www.informationisbeautiful.net/visualizations/million-lines-of-code/

 

46 minutes ago, RedRound2 said:

Again, I am not talking about some complex system with over 1000 different data points. It's a damn ECU that an Arduino could do.

Except that it is a complex system with over 1000 different data points. An arduino maybe could do the job of an older ECU in an older, simpler car, but I doubt it would be able to replace anything other than the most dumb ECUs found in modern cars.

[GDRRiley]

18 hours ago, Kisai said:

Power Steering has been on ALL vehicles since the 90's, It's simply impossible to drive a vehicle that doesn't have it because the amount of muscle strength needed to turn at low speeds is too high for most people who didn't grow up on it.

I'm use to non powered steering. Its not that bad and I'm 20

 

for some products it doesn't make senses to shrink as you gets no real benefits.

but when it comes to infotainment and the such yeah they need to update the back end.
I get why though they struggle to update the computer parts on the mechanical side. I assume its mostly 10 year product cycles

[AnonymousGuy]

Shortage is probably intentional to a degree because  covid had nothing to do with slowing down semiconductor manufacturing.  If anything it was the only operation that was "at all cost" and kept running even as everything else around the fab was emptied out.  So now you have a sweet excuse to jack up your pricing on your 30 year old fab that was probably barely worth operating pre-pandemic.

[JForce]

23 hours ago, Zodiark1593 said:

I kind of disagree. Reliability really trumps just about everything else when it comes to vehicles, especially as a number of functions are being controlled by the infotainment system. 
 

If vehicle makers can move to newer processes in a timely manner without compromising reliability (which will take time and tons of validation), then by all means do so. However, rushing over to new silicon is not the way to go. And with drive-by-wire slowly coming into play, failures are really not an option. 

I don't think anyone, including the chip and car manufacturers, disagree with you. I think the "timely manner" part is the crux of it.....many of the electronic designs in your car are at least a quarter of a century old, and I think the overall point is that they're now facing the other side of the "long tail of reliability" coin.

[CarlBar]

3 hours ago, RedRound2 said:

Is there any proof for this 4 year or more claim? Coz I know for a fact Tesla doesn't and they're very open about what the cutting edge hardware they are using. And I highly doubt anyone else does either. Unless you're talking about some super complex system (which honestly im not sure what such complicated system is there in a car apart from self driving) testing ECUs aren't such a huge deal that people are making it out to be. 

 

I'd have to dig around on the specifics. but warranty reasons alone demand a 4-5 year proofing to ensure they meet that. if your not customers and regulatory agencies will take a big dump on you. Note when hardware comes out has zero bearing on when testing and validation started. Tesla AFAIK is also more willing to do short tests so they can get it out the door, then continue that over a longer period and if trouble shows do recalls.

 

Also to add to the later stuff. When your testing long term durability the only way to do it is to test for a long time. No way around that. Being a multi-billion dollar company with huge numbers of people can't make things that need time happen fast.

 

4 hours ago, Dracarris said:

However, many early planes with mechanically controlled control surfaces use hydraulics only for assistance (like power steering in a car). Should hydraulics fail, you can still use (a lot of) muscle force to actuate control surfaces as long as you are not in a flight path where aerodynamic forces are so high that it's not humanly viable (e.g., in the B737 Max erroneously commanded heavy nose down).

 

For military fly by wire systems in fast jets it's generally because they're a reduced or zero stability design, the aircraft is just flat out unflyable without computer assistance. the 737 MAX under some conditions is the same.

 

For civil use it's a case of jet airliners have long since got so big and fast you can't do it via muscle power alone. There's complete loss of control accidents going back to at least the late 80's, (on a DC-10), due to complete hydraulic failure.

[Donut417]

39 minutes ago, AnonymousGuy said:

Shortage is probably intentional to a degree

Nope. TSMC was building a new fab in the US even before this shit went down. Company like TSMC has projections for demand. With how much more tech we have year after year they knew demand would be growing. 

 

The only reason the auto makers, at least in the US are in dire straights was because they told chip makers during the beginning of all of this shit that they DID NOT need the capacity. Because they thought demand was going to go down. How were they suppose to know that the government was going to give away free money? And everyone instead of saving it was going to go out and buy a new car. 

 

Furthermore, if covid didnt hit, we likely wouldn't have had a supply issue like this. Because generally there is a surplus of goods (insert product here). This way if there are production issues, you wouldn't see wide spread shortages. The problem was, between the droughts in Taiwan (TSMC has a Fabs there), the lock downs and shut downs of businesses around the world. All of the excess supply was sold and there was little to no new stuff being manufactured for months. Keep in mind, China shut down, China is the industrial capital of the world, much of the world relies on their factories. 

 

The part of this no one seems to be discussing is the supply chain issues. There are shortages on shipping containers and ships. My employer used to be able to get a container and a spot on a ship for about $4,000 I was told. Now its $20,000 and no guaranteed spots on a ship. We have 14 Warehouses spread across the US. So generally an order is shipped from the closes warehouse. The Detroit warehouse where I work has been shipping to California, Texas and South Dakota. California and Texas both have warehouses, but they dont have the product. South Dakota where where our HQ is based, they dont have product either. I guess as soon as a shipping container arrives the product basically gets received in to inventory and goes straight out. 

 

[CarlBar]

@AnonymousGuy to add to what was allready said the stay at home nature of the pandemic sent electronics purchases through the roof. The supply chain has some ability to cope with jumps above expected. But unless they're short term, it can't keep up until new production capacity comes online. And that takes years in semiconductor manufacturing.Tats the downside of "Just In Time" supply chains. they don't cope with sudden upswings well, but handle downswings much better, (less excess inventory sat around). And the latter is valued because downswings can kill companies at the best of times. Too much unsalable stock makes them worse so making them as low pain as possibble is desirable. 

[RedRound2]

8 hours ago, leadeater said:

Yes and that includes the ECU if you update it to a new model not used and gone through the testing. And no an Arduino could not do it, it'll immediately fail environmental test conditions, again complexity of the task aka the code or compute power isn't really a factor it's making sure it actually does it ALL the time, as best as possible.

 

I would never put my life in the hands of a Arduino control car ever.

You always completely miss the point. I brought the comparison of an Arduino because the task an ECU is supposed to do is more or less very basic in a low to mid end car. That means the software itself is quite simple to make it not prone to corner cases and error. Envrionment hardening is done at manufacturing so it's pretty much already applied and very less likely to go wrong. The only thing to be tested is the new processor node itself, whether it works relaibly or not. And given that the same sort of architectures and chips are likely to be used, we already sort of have data for it from the vendor itself like Nvidia, or AMD or Intel, or ARM

8 hours ago, leadeater said:

That doesn't change how much testing is required or the complexity of testing. Testing something to the safety standards of vehicle regulations I would say is more complex than designing the ECU in the first place, if not more complex at least more time consuming (person FTE equivalent).

 

You're saying it's not a complicated system to test and your reasoning is these are huge companies with many employees and billions of dollars, all facts that have nothing to do with testing complexity.

 

Telsa has never rapidly changed computers. HW1 was Tesla designed hardware and software (2014-2016), HW2 was Nvidia designed hardware (PX 2) Tesla software (2016-2017), HW2.5 was Nvidia PX 2 again with Tesla software changes, HW3 Tesla designed hardware and software (2019-).

 

I would hardly call any of this rapid when they have so few models of cars and do not update the car platforms in the same manor as other manufacturers.

Oh so they only care about hardware changes (that by default is usually thoroughly tested unless the manufacturer makes their own architecture - which they dont) but not the frequent software updates that can easily introduce bugs into the code? The logic seems so stupid tbh. There are standards and guidelines for enviroment protections and tolerances - but that it's. Nobody goes inspecting of any of these things.

8 hours ago, leadeater said:

No the answer is car platform update lifecycles. Traditional manufacturers do not change hardware or software on existing in production cars, Tesla does. So if you are buying a 2019 year model car that is a 2015 platform model then you have 2015 technology in it not 2019. Tesla has and started with even less models of cars and also restricted which markets they sold in to lower the burden of regulation compliance, an actually smart thing to do.

And all im saying is replace that 13-950 with an i3-10300 or whatever it is.

8 hours ago, leadeater said:

But may I remind you I was replying to you about your comment that testing ECU isn't complex, which it is, and there is a ton of regulation compliance required for vehicles which covers this and the actual business as a whole (IATF 16949 as one example). You think it's simple because the function tasks being carried out is simple, that's just an assumption you've made.

 

Vehicle design and safety is literally life and death and somehow you think the testing required for components like the ECU isn't complex an laborious? I work in the public sector, if something can be made complex and nonsensical they'll do that and double it.

It is quite simple compared to most things we do today. A full platform update like how Tesla does is the one that is scary. Even though they do it few years apart, that computer controls literally everything in the car. ECUs don't. And they're usually always with mechanical fallback system because we still dont have a complete drive by wire car

8 hours ago, igormp said:

Those tiny processors nowadays are pretty powerful by themselves, since we already have a wide amount of dual core microcontrollers out there with caches and whatnot.

Along with that, you also need integration tests since even an entry level car has over 40 ECUs and more than 5 different communication buses. High end ones easily have over 100 ECUs and 10 buses.

 

This is a little outdated, but gives a rough idea how large a car codebase is: https://www.informationisbeautiful.net/visualizations/million-lines-of-code/

 

Except that it is a complex system with over 1000 different data points. An arduino maybe could do the job of an older ECU in an older, simpler car, but I doubt it would be able to replace anything other than the most dumb ECUs found in modern cars.

I'm not talking about revamping the code. I never was. The code you write should be typically not for one specific processor. I've said all I wanted to say in the above lines.

 

All of you are justifying car manufacturers with shitty systems, but you know ever since Tesla entered the race and lit a fire under their butts about acutally having a good EV and software experience, these people are starting to come out with cars with massively improved systems running on the roads today that they definitely have not tested for more than a year at max). And surely they can pluck parts of those system - like the infortainment and put it their regular lineup as well - but nope

[leadeater]

7 hours ago, RedRound2 said:

You always completely miss the point. That means the software itself is quite simple to make it not prone to corner cases and error.

No it doesn't, that's just another assumption. You do know the all cars, even basic ones have O2 sensors, sometimes EGT sensors, and fuel flow and fuel type adjustments that happen in real time. Also spark timing as well, crank position sensors, valve timing and position. Other stuff I've either forgotten or don't even know about.

 

Even if we go with the assumption an Arduino could run in a sterile lab type environment you have no idea in what way it would fault and create senor read errors at high temperatures, low temperatures or shock vibration and just because you've got some basic error handling code does not mean when the hardware itself it faulting that it'll get handled correctly.

 

I think you missed the point not me, Arduino could not do it because it would never pass environmental testing so it could not be used to control anything in a car, an engine on an engine stand maybe and that would likely be without all the automatic fuel and compression adjustments.

 

7 hours ago, RedRound2 said:

There are standards and guidelines for enviroment protections and tolerances - but that it's. Nobody goes inspecting of any of these things.

Wrong. I like how you are so sure of yourself that you didn't even bother to look at what IATF 16949 is.

 

7 hours ago, RedRound2 said:

And all im saying is replace that 13-950 with an i3-10300 or whatever it is.

Do you even know what any of the computer systems in cars look like and how they are designed?

 

7 hours ago, RedRound2 said:

A full platform update like how Tesla does is the one that is scary

Tesla has almost never done a full platform update of any of their vehicles, they introduced knew ones. I don't think you know what a vehicle platform is.

 

BMW E30, E36, E46, E93, those are vehicle platforms. How many of these has Tesla done for any of their models of cars?

 

7 hours ago, RedRound2 said:

And they're usually always with mechanical fallback system because we still dont have a complete drive by wire car

We have extremely close to that, only breaks are not this. We have electric throttle control and electric power steering. Electric power steering uses torque input from your steering inputs and relays this to an electric motor, your steering wheel output is not directly connecting to the steering rack anymore which is where the primary complaints from lack of car feel come from. Further to this many types of electric power steering systems have dynamic input adjustment at different driving speeds so the turn rate is not a fixed amount. While there is fallback to manual steering wheel control there is still the danger that the electric motors not operating correctly or doing erroneous steering control.

 

There is no mechanical fall back for electronic throttle control however.

 

Your post is full of so many assumptions and so little understanding. Electric motor control is actually simpler than a combustion engine as well, comparing the engine management between the two isn't even worth the effort.

 

I don't even know why you think I'm even slightly anti Tesla, just addressing your wildly misinformed assumptions about ECU complexity and testing. There's really good reason manufacturers use the same ECU across multiple car platform updates because there can often be no reason to update that component. Toyota does this the most, Honda as well and which two car brands are most know for reliability? I wonder why that is.

[CarlBar]

3 hours ago, RedRound2 said:

There are standards and guidelines for enviroment protections and tolerances - but that it's. Nobody goes inspecting of any of these things.

 

Kind of irrelevant point. It is in no way not acceptable for the companies to meet those protections and tolerances even if no one's checking, (unless of course something goes wrong and then they will check in excruciating detail).

 

3 hours ago, RedRound2 said:

And all im saying is replace that 13-950 with an i3-10300 or whatever it is.

 

An i3-10300 doesn't meet the temperature range spec, let alone the rest of it. And thats before we get into packaging it to meet all of the vibration, shock loading, grime exposure, dirty power, RF noise, e.t.c. factors. one of thats trivial or easy. Could it be done. Sure. But it's not quick or cheap to do. it takes time and effort to do.

[Dracarris]

2 hours ago, leadeater said:

While there is fallback to manual steering wheel control

How so if the steering wheel has no mechanical connection the wheels anymore?

[leadeater]

47 minutes ago, Dracarris said:

How so if the steering wheel has no mechanical connection the wheels anymore?

There is a direct connection however the actual output from your control under normal operation is relayed through a motor and output as a computer controlled calculation. This is either done through an angled worm gear or a belt connection, basically the steering wheel to steering rack is a lose connection until the electric assist fails then it's a tight connection.

 

 

This is one type of electric power steering, the motor is attached via a belt to the steering control rack and rotates it. The steering pinion has a torque sensor (the piece around it with the control wire coming out) and that is an input in to the steering control motor. If you were to lift the car and apply power to that motor the steering wheel in the car would turn with it.

 

The reason I say it is a lose connection is that steering pinion is not applying direct rotational force to the steering rack, the electric motor is. Maybe floppy or gear play is a better description of the effect.

 

Here is another type

 

There is quite a few different designs with different levels of control and feedback feel. I think this above one is the type with more feedback feel than the first.