f13.net

Surprised this hasn't been chatted about...

http://www.bbc.com/news/business-34373637 (http://www.bbc.com/news/business-34373637)

I haven't seen anything about the Audi and BMW divisions, and so I have no idea if this affects them. Mass quantities of money here, both in labor, and parts, as well as stock and what I am sure will be punitive damages. I'm sure VW can stand through this, but, I'm not so sure, ya know?

VW doesn't own BMW.

VW shares engines and engine management systems across all brands so this issue likely affects all VW brands. BMW is not a VW division though but a wholly separate company.

VW holding owns:

Passenger Cars:

VW
Audi
Porsche
Skoda
Seat
Bugatti
Bentley
Lamborghini

Trucks:
Saab/Scania
MAN
Neoplan
VW branded

Motorcycles
Ducatin (wholly Audio owned subisdy)

BMW and Daimler may not have cheated on that particular EPA test cycle though. The test VW has cheated on involves limits on Nitrous Oxide levels in the exhaust of diesel engine cars.

There are several ways in which you can conform to those regulations. The best one involves urea injection systems (brand name AdBlue) but there are other ways to reduce NOx levels as well. Those systems add to the cost of the vehicle though (additional tank, injection pump and sensors). The additional cost is an issue for cheap entry level cars.

That's why most competitors have either stopped selling diesel as an option for their cheaper models in the US or they use urea injection or similar systems to comply (BMW or Daimler for example). VW chose the 'cheat on the emission testing' option instead.

This could not have been accomplished without the knowledge of VW executives, regardless of what they claim. For one because due to VW's unique structure the CEO level of executives is still involved in that sort of strategic technical decision-making and also because Winterkorn and indeed most of the executives fired recently have at one point been vice presidents of research and development or even vice presidents of engine development.

Lenny Kravitz!

Can I get a tl;dr on what's going on? I know fuckall about cars, so the article didn't make much sense and the thread title doesn't help. VW lied about their cars emissions...or something?

VW programmed the computers in their diesel cars to detect when they were being measured for emissions, whereupon the car would essentially change into a new mode where it emitted way less pollutant to as to fool the test. On the regular road, not in a test, the normal computer control over fuel and emissions took over and the cars emitted way, way more pollutants. Which was VW's way of avoiding more expensive designs.

So the people who thought they were buying a very environmentally responsible diesel were actually buying cars that violated existing environmental standards. Which means the cars just instantly lost much or all of their resale value.

Plus violating laws and committing fraud. It's a pretty big deal all around.

To add on:

They were caught when a company made up of ex-EPA people went through the difficultly of checking the emissions from the exhaust pipe(s) while the car was operating on the road*. VW spent more than a year denying they were lying cheating scumbags until the evidence was too great and they finally admitted they were lying cheating scumbags.

* it's physically difficult to get all the equipment hooked up properly to a moving vehicle which is one of the reasons why cars aren't tested like this normally

Edit: Also the EPA is actively blocking researchers trying to get an exemption to the DMCA so that they can inspect the code in on-board computers for these kinds of shenanigans.

Good thing the only people that buy VW's are basic bitches. ಠ‿ಠ

Fuck you!

Now, who wants Starbucks?

For the curious on the DMCA-blocking part: http://www.slate.com/articles/technology/future_tense/2015/09/volkswagen_s_cheating_emissions_software_and_the_threat_of_black_boxes.html

Wow, that IS pretty fucked up. Thanks.

Honestly, it's nothing new. The pearl clutching from people over this is hilarious. Pontiac cheated on their EPA test back in 1973 doing almost the exact same thing (relay timed exactly for the test that shut off emissions equipment after).  Their excuse was "you didn't say it wasn't ok to cheat" which changed the rules. The big truck diesel engine manufacturers got caught cheating in the late 90's & early 2000's, which is why Ford lost their 7.3 engine and got stuck with the 6.0 - changing emissions requirements. Caterpillar recently got out of the diesel engine vehicle business because of the EPA requirements. It's a mess publicity-wise for VW, I think what's worse is that it went on so long & they didn't get caught. The EPA requirements for emissions aren't exactly the most logical, they push tailpipe emissions over fuel economy & driveability - so it's better to lose 20% fuel economy for a slightly cleaner tail pipe emission. The running joke is they're only satisfied if your car pulls in dirty air and emits clean air you can breath directly.

People who buy a car because it's more "ethical" or "green" compared to others are just buying into marketing hype and those types tend to get what they deserve.

Detailed replies to that would take it to Politics. Let's just say that I don't see what the diff between any of that and Kim Davis saying she doesn't like the law on gay marriages is, if the upshot is, "Hey, laws are now totally voluntary if you think they're silly/immoral/irrational".

For those not interested in debating it, the facts are now pretty clear, at any rate. And it turns out by the way that VW was doing on European cars too.  And as per the Slate article, VW is one of many companies fighting quite seriously to make it illegal to look inside their computers in cars. I wonder why.

Turns out I should have just watched John Oliver last night instead of waiting until today. :why_so_serious:

No argument from me at all there, the rules are the rules. I was just pointing out that companies have been cheating and getting caught (or not!) ever since the EPA started testing 43+ years ago. And that everyone acting like this is something new or unheard of amuses me ;-)

GM has been doing this a good while now, even to the point of trying to make it illegal to buy parts or computers not from them (and require that you have to use them for firmware upgrades, etc). So far they haven't succeeded beyond voiding your warranty if they suspect you changed something.

There are several problems that make this go *way* beyond "VW cheated on EPA tests, stupid greenie tree huggers might as well be rolling coal."

1) The resale value of a diesel VW is now approximately nothing. It will probably affect the value of all VW's (I drive a non-diesel Jetta).

2) In several states, driving an affected VW now that you are aware of its noncompliance may be illegal. In others, it just can't be registered or have the registration renewed.

3) VW got caught because they cheated so hard that it made engineers say "WTF?" It just wasn't physically possible for an engine with that design to perform that well and be that clean.

Which raises the question: How many other cars were they doing this with, in a less extreme fashion? How many other manufacturers are doing something like it?

--Dave

GM has been arguing that you get a license to the car when you "buy" the car from them (http://consumerist.com/2015/05/20/gm-that-car-you-bought-were-really-the-ones-who-own-it/), and that means you don't actually own the car. Among other things, they could make the license non-transferable (can't resell the car), or place all kinds of restrictions on use.

(technically, you might own the hardware of the car, just not the software without which the car cannot operate at all; it'd be like when you "buy" physical software all you own is the physical media of the CDROM/floppies/etc but have no ownership rights to the bits on that media)

At least this shows how bad the DMCA is and how easily it can be used to fuck over customers. In industries that have nothing to do with digital media.

By the way, I hate the label the press gave this trickery. It is not a defeating device it's much simpler than that. People that claim that this is some sort of elaborate cheating software - maybe even explicitely designed by the supplier of the engine control unit - are missing the point and show that they have no clue how the business works.

Each and every car that uses electronics to control part or all of the vehicle basic functions has a test or dyno mode. Explicitely for performance and emission testing done by authorities or for road and safety certification testing. Basically if you do any sort of testing on a dynamometer or on a stationary vehicle you have to put the vehicle into a special dyno-mode or most of the safety-related functionality (i.e. traction control, stabilitry control, antilock braking, power steering, engine control) will freak out completely. You know as they are supposed to do when only two out of the four wheels are turning and there is no steering input. There is a whole set of requirements documents for electronic control units that only governs the dyno mode and how it needs to be implemented.

This is not an elaborate cheat SW, they basically only detected that the vehicle was being tested and then switched to a different engine setup. Which is - minus the switching to an EPA-safe mode - exactly what the system is supposed to do. This is trivial even without someone entering the dedicated test mode since all necessary data is exposed on the vehicle's internal communication network. It is such a basic trick that I wonder what led VW execs to believe that they could get away with it. It's basically using a documented and commonly used emission testing mode to cheat. This had to be discovered eventually by anyone that is not totally naive and braindead.

It's also probably not a deliberate engine setup for cheating explicitely made by the supplier for that purpose.

Firstly because there are a lot of legitimate uses for test modes and non-standard SW setups during development, i.e. while the power train is being developed and tested outside of a production car environment. Secondly because the letter from Bosch many sources quoted contains merely standard legalese you always include if you provide prototype or testing SW to an OEM. We use similar standardised phrases when we supply our customers with pre-production SW simply to not be held liable if a pre-production build with test and debug modes enabled somehow ends up in a production car (which can and has happened). Lastly because the automotive business doesn't work like IT. The supplier doesn't deliver a pre-compiled and finished SW binary that the OEM only installs on the ECU or a pre-packaged ECU that is only plugged in and attached to the network. VW is developing the engines not Bosch. They have in-house developers that extend and modify the code and adapt it to their different powertrains and can and do make quite extensive modifications to the (usually source level) code supplied by the partner.

We usually get all of our ECU SW modules that we contract out or buy from third parties as source code - even modules that include proprietary and patented code - that we have to adapt to some degree. The certification requirements necessary to put the SW into a production car basically require source level access. It usually comes with several dozen pages of legalese that can be boiled down to "the SW is certified for an intended use a,b and c and if you make any modifications or operate it in modes that deviate from intended use then everything is on you and we won't assume any sort of liability. Good luck."

I'm not saying that this might not also be inteded as ass-covering by the supplier. The relationship between suppliers and car manufacturers is one where they make you an offer you simply can't refuse without potentially going out of business. I won't say more on that topic but given that there are only a few big corporations that all manage purchasing decisions globally for all sub-brands and only offer pretend choice due to the dozens of brands they have, pissing off the second largest car maker after Toyota is not a winning move if you want to stay in business.

you care too much

It's a German company affecting Germany exports and individuals. I'd be more surprised if he DIDN'T write a few pages on it. That's just Jeff.

It'd be like a big report coming out saying "Vaping is actually more toxic than smoking: here's how they hid the results" and you didn't comment.

Good for him for caring. If you don't care, you deserve what you get.

America gets dragged through the mud every day and I don't give a shit. ¯\_(ツ)_/¯

Patriotism is for my grandpa's generation. I'm a galaxyman.

I don't get the vibe that Jeff is motivated by some kind of Landsorgen here. He's not mentioned a German angle at all. He obviously has some experience of how these things work in his professional life and is sharing it to clarify why stuff like the test mode exists and why it's vanishingly unlikely that some rogue engineer did this without the explicit knowledge of very senior people.

Thank you.

Take the now Ex-CEO of the Volkswagen Group Martin Winterkorn.

He started his career at Bosch - the supplier of the engine control systems for the affected Diesel engines in the eighties. He went on to become VP of quality assurance at VW (brand) in 1993, VP of product development at VW (brand) in 1996 and VP of R&D at VW (brand) in 1997. He then went on to become CTO/VP of R&D for the VW Group (all brands) in 2000 while also holding the title of VP of R&D at Audi in 2003 and he has been VW Group's CEO since 2006. He is well known as a person that was still micromanaging technological decision making and product branding at VW when he was CEO.

VW has now fired a dozen managers, all of which have at some point been VP of R&D or VP of engine development for one or more of the VW brands, all of them have been reporting to Winterkorn since at least 2003 when he became head of R&D for all VW brands. There is no way in hell that he didn't know about and authorize all of the cheating that now came to light.

There is no way that this wasn't known from the lower levels of engineering all the way to the top. You don't do something like this without everyone in the food chain knowing.

I find this all to be very strange. In NJ, whenever I had to get my cars inspected, I went to a DMV facility. It would appear that, to test the emissions, they hooked something up to the pipe (maybe the onboard electronics as well), and essentially ran the car for a few moments on what was essentially a dyno.

It's the "maybe the onboard electronics" part that's important. If you car has an on-board computer they will connect to the diagnostic port as part of the smog check protocol.

I don't think there are any cars made now without an OBD-II connector, although frequently it is not obvious (generally under the dash, somewhere near the steering column). If you've ever gone to a mechanic because your "Check Engine" light was on, this was the first thing they did (hook up a reader and check the codes).

At any rate, merely being *on* the dynometer is enough different from normal driving that the ECM can detect that it is being tested, even if it is not explicitly going into a "test" mode from OBD instructions, if the car has any kind of traction control system (and I think all VW's do). This was noted above. There was nothing particularly challenging on the technical side to produce this cheat.

--Dave

Edit: It's worth buying a reader and finding your ports, just to keep your mechanic honest. The prior owners of one of my ex-wife's vehicles were soaked for thousands in repairs, when actually a malfunctioning sensor was occasionally and randomly reporting that *all* the cylinders were misfiring. The sensor just wasn't always turning off fast enough when the engine was shut down, and there's nothing wrong with the engine at all.

Love to see the commits on those builds.  I'm sure there's a single long German word for the cognitive dissonance.

QA:  "Thing we can't talk about is not working as designed, which is correct.  It is working non-functionally."

Sorry, Soln. Since it has a common Latin source, its just "Kognitive Dissonanz".

Regarding every engineer that worked on it knowing about it - I doubt that. The higher-ups surely, but not every cog in a machine knows what the other cogs are doing. And if they were smart they divided the software code and the "switch into test mode" people and the "change emmision settings" people didn't know how exactly their modules interact.

Oh fuck, please don't. Terrible, terrible advice.  If you don't trust your mechanic, then shop until you find one you can. If you bring your car to me, and you've erased the codes on it because you didn't know what you were doing, you're gonna have a bad time. Learn about cars or don't. Don't halfass it thinking you're going to be able to outsmart the automaker and the guy who has spent thousands of dollars and hours to learn the trade.

Why are all mechanics posting on the internet always honest and up front and all the ones in real life slightly worse than Dr. Mengele, Auto Surgeon.

I had a long post, but the internet ate it. I'll give the short form: Code readers don't erase codes automatically or spontaneously, and if a mechanic had a problem with me reading the codes, I would solve the problem the same way I treat anyone that I pay that prefers his clients to be ignorant.

--Dave

I'm well aware how code readers work. And we'll count ourselves grateful that you're not my customer.

Because the crooked ones are busy working and attracting clients while the competent ones are on the internet giving shit away for free?

This is one of those things where having a "guy" for everything helps. I've got close family friends who basically cover every single possible thing I could ever need repaired since I still live within close driving distance of my hometown. The mechanics I know (general, a body shop, and a transmission shop) are all old war buddies of my dad so I pay cost for quality work whenever I need it done.

No you'd probably spend three times the amount of money on another mechanic to even find out what's wrong.

Those diagnostic trouble codes are about the only thing a mechanic has today to help him isolate a potential problem easily and without a lengthy process of isolating the fault. If they get deleted or altered it will turn the fact that for example the engine control module is complaining about a defective lambda sonde (a common fault) into a lengthy search as to why your exhaust management system is shitting all over itself and ruins the catalytic converter of your car in the process.

Using an aftermarket OBD or OEM diagnostic system when you don't know what you're doing can break your car in ways not even an officially licensed and authorized shop will be able to fix.

The insane competition between brands in an overcrowded market and the way the car companies have set up their dealer networks and designed the contracts means that officially authorized dealers don't make any money on car sales. They'll likely even lose money on the initial sale if they are operating in for example Europe which is a crowded and highly competitive market. So they make their money on mandatory servicing required by your car company's warranty and aftersales service and repairs.

Since the car companies have to rely on their official dealer network for sales they'll do as much as they are legally allowed to do to 'help' official retailers and authorized resellers make money on after market repairs and servicing. By for example making it so you need testing and debug equipment to even be able to see what's wrong with a car. Equipment that is only being supplied to authorized retailers of course. Or by making it so only authorized retailers can buy spare parts or by using proprietary fixings and only making the tools available to authorized shops. Or by voiding your warranty if you're not using 'authorized' motor oil or other lubricants. You get the idea.

You are more likely to get an inflated repair or servicing bill at an authorized shop because that is the only way most of those shops actually make money.

Would they throw out a bunch of FUD about OBD readers, making it seem like the most dangerous thing you could do is hook one up, spewing technobabble like "lambda sonde" instead of just saying "oxygen sensor"?

I'm not talking about trying to reprogram the ECM (or even trying to do your own repairs), but simply finding out what the actual problem is according to the diagnostics before going to a mechanic. A code reader is a passive instrument, the only thing it can do is read the codes, and clear them if you *actively* instruct it to.

Even if you do (which yes, you shouldn't if you don't know what you are doing), if there's an underlying problem the code will come back and the CEL will turn on again.

--Dave

No it's not FUD.

The problem is that the vehicle network bus access port is vendor specific and proprietary and the documentation is not readily available. Implementing a 3rd party interface according to manufacturer spec is virtually impossible if you are not licensed and don't want to breach NDA's or flat out ignore copyright and IP laws. Basically each car manufacturer has a few licensed vendors of equipment that only sell to manufacturer authorized companies and retailers (and need to verify eligibility for sale with the car manufacturer). So the only entities that have legal access to the tools and the documentation are authorized retailers and shops.

So all aftermarket readers rely on reverse engineered information or the small portion of the specs that are publicly documented because they are part of an ISO standard (like OBD II). You're also dealing with people that don't really like you - or more importantly your friendly 3rd party neighborhood mechanic - being able to read out vehicle maintenance info from your car becauise that threatens their business model.

There is a real risk that shit can break when you use aftermarket debug interfaces on your car. In ways that even an authorized shop can't fix because it concerns settings and data only the car manufacturer has access to or can rewrite.

Probably not when you only read out Diagnostic Trouble Codes. That's not why aftermarket interfaces are sold though.

VW for example is so restrictive that none of their authorized retailers for one of their sub-brands is allowed to or even physically able to repair a car from another sub-brand. So if you roll into an Audi shop with a VW they might not be able to service your car. Even though VW extensively shares components across vehicle platforms and brands and has lots and lots of component reuse.

Component reuse is what made this fraud affect 11 million cars in the first place. Audi A3, Audi TT, VW Golf, Skoda Fabia, Seat Leon, VW Scirocco and countless others. Pick your favorite brand. Congratulations you're still essentially driving a VW Golf. Don't like the Porsche Cayenne? Great buy an Audi Q7, same basic platform. Yet an Audi dealer would not be able to fix a VW Golf even though he basically sells one rebadged as Audi A3 and fitted with a few additional niceties.

Having called a Nissan dealer about my Infiniti G37, which is a Nissan Skyline in Japan, I can corroborate.  I also paid +$500 for two key fobs, so again I corroborate.

Which is why I have repeatedly used the term "Code Reader". Yes, anything that will let you mess with your control electronics is inherently risky, but a simple OBD II code reader is restricted to *reading codes*. The good ones have a table of what the codes correspond to in plain english.

All I am saying is that the CEL is an 'idiot light', it just tells you the on-board diagnostics have detected...something. If you don't know what that something is, you're at the mercy of your mechanic. I have one of these (https://www.pricefalls.com/product/obdii-auto-scanner/58782820?source=GoogleShopping&medium=cpc&term=PLA-SC&content=131081086591product_type_l1home+improvement+%26+tools%26custom0p4100200%26id&campaign=CoastalToolSupply&adpos=1o9&creative=62717849431&device=c&matchtype=&network=g&gclid=CLTFnvGersgCFYNFaQod-V4D0A) (outdated model, bought it back in 2010), please tell me how it can break my car.

--Dave

My Hyundai automatically cleared codes any time they were read (even if it was a "read only" code reader).

Ha, suckers.  I've avoided all such worries by simply not owning a car for the last 10 years.   :awesome_for_real:

Just a heads up.

The way the OBD (on board diagnostics, ISO/DIS 15031), KWP 2000 (Keyword protocol 2000, ISO/DIS 14230) and UDS (unified diagnostic services for CAN, LIN, Flexray and MOST based systems, ISO/DIS 14229 and ISO/DIS 15765 for the CAN related implementation) are designed and implemented makes it so that technically no device can be a pure read-only device.

Even just to gain access to fault memory and DTC info, write level access is required. The vehicle also registers diagnostic access with timestamp and additional info. Some implementations of fault memory and DTC also automatically clear codes once read. The shop can usually see that somebody else has accessed the diagnostic port and used the interface to do something if they bother to look for it.

It will also void your warranty if you care about such things and might even make official shops refuse to service your car to not assume liability for the things you might have fucked by accessing the debug port.

[edit: unfucked a sentence and added more info]

This thread has gotten stupid.

FBI arrested former regulatory compliance chief in the US, Oliver Schmidt, on Monday.

https://www.engadget.com/2017/01/09/report-fbi-arrests-volkswagen-executive-over-dieselgate/