Volkswagen Is Ordered to Recall Nearly 500k Vehicles Over Emissions Software | Hacker News Reader

Volkswagen Is Ordered to Recall Nearly 500k Vehicles Over Emissions Software

nytimes.com

353 pointsAmorymeltzer323 comments

I research for the auto industry and this was very well known by the authorities. Fixed test cycles like the New European Driving Cycle https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Ne... were easy to detect, and easy to optimize against. That's why coming regulations introduce more realistic cycles like WLTC https://upload.wikimedia.org/wikipedia/commons/thumb/b/be/WL... and real road tests, where particle emissions will be trapped and compared against the lab test, and only a certain % of deviation will be allowed (last time I looked they didn't decided this % yet).

I hope this gets fixed, although I don't believe it. I care about the noise emissions. AFAIK they are measured on a real road, but only up to 50 km/h and in a ridiculous narrow test.

Harley Davidson and other are known to optimise for this test with movable parts in the exhaust pipes so these get much louder when the bikes goes faster than 50 Km/h.

I actually work on aeroacoustics! My specialty is turbomachinery but regarding combustion noise, let me say that it is a very tricky subject... even slight variations on the fuel injection settings and valve openings can have a high impact on noise!

Well, as much as a cheat it is, I'm happy that those bikes can go quiet on slow (city) streets but as loud as possible on the highway (increase security and swag).

Loud Harleys frustrate and annoy me. Most owners seem to remove the factory pipes and fit the loudest ones possible.

It's not true this helps on the highway. In my car, which has double glazed glass, you can't hear a bit until it's right outside the window.

When I'm walking along with the dog and kids, an open pipe Harley riding past is particularly not safe, because it scares the crap out of them.

One idling Harley is louder than a thousand idling Volkswagen diesels.

I'm a motorcyclist too, and straight piped Harleys make me stabby. I don't believe they're louder on the highway. I believe they're louder whenever they're within earshot of any place where people congregate without earplugs (cafes, restaurants, pedestrian areas), whereupon the scumbag on the seat finds it necessary to go wide-open throttle to show off how terrible their bike sounds. It never fails.

I'm always surprised when people don't realize that large bikes are suppose to be loud so that drivers can hear them coming. It's a necessary evil.

I'm a motorcyclist and I'm downvoting this fallacy. If this were seriously the case then good beginner bikes like the Ninja 250 and Honda CBR250R would be unsafe, especially for new riders. Enjoy your loud pipes if you like (I do on my throaty Aprilia) but don't pretend they give you any safety benefit to balance the annoyance it causes some people.

I too am a motorcyclist (85K miles of sport bike riding, Sunday morning ride to Pt Reyes was a lot of that, current stable is r1200gs and wr250r).

I despise loud pipes, just frigging hate them. But I live in deer country (Santa Cruz mountains) and I've hit a deer at speed on the GS (did not wipe out, lifted the front end and had my front tire hit it square in the hind quarters, spun the deer around, I dropped the front wheel down and did a very shaky coast to a stop. My hands were shaking like a vibrator on steroids).

After that experience I really wonder if loud pipes would have scared the deer off. I tend to doubt it, I've not installed a louder pipe, but it's made me less pissed off at the loud pipe people. Not a lot less pissed off, but some.

BTW, shoutout to the Ninja 250 and especially the 500 version. I had two of the EX500's and those were by far my favorite sport bikes. Lots of torque, could pull any 600rr off the light (yeah, the 600 was blasting by me 2/3rds of the way down the block, but do you really want to ride a sewing machine? I don't. Loved my 500 twins.)

+1 here. I have two bikes: one is a custom tracker (read: bunch of parts bolted together into something that runs... mostly) with open pipes. The other is a Zero SR (electric). Although anecdotal, I notice no difference in driver perception of my presence, except in stand still traffic, where it is least needed.

With the level of acoustic engineering that goes into modern vehicles specifically to block out external noise, combined with radio / in car entertainment, relying on other drivers being able to position you from audio is a bad idea. Now if bikes were to be fitted with cell phone jammers, that I could get on board with.

TL;DR - I stand by my comment, loud bikes helps people hear you but that doesn't mean you have to be EXTRA LOUD to be cool.

I understand the point you are trying to make, and perhaps I wasn't very clear in my off-hand comment. I agree that overly loud bikes are annoying, however, as I stated before, sound is a necessary evil[0].

Do I believe that motorcycles should use straight pipes and increase their sound output? NO, in fact its quite annoying. However, louder bikes will get your attention.

In fact, in reference to [0], there is a standard that some vehicles had to INCREASE their decibel output because of less risk to pedestrians. This isn't an apples-to-apples comparison as this study was mainly geared towards electric vehicles, but it did state that vehicles should have an output of ~ 55dB for pedestrian safety (page 102).

Secondly, I couldn't find good stats on motorcycle decibel output, but I did find a study of some motorcycles and their average output[1]. What I learned is that most bikes are fairly equal with regard to output and all are at, or above the, legal limit[2].

I took a single bike you pointed out `Honda CBR250R` and was curious how the noise output performed to a Harley: `Honda CBR250R`: [3] - With dB-Killer fitted: 86 dB @ 4,250 rpm - Without dB-Killer fitted: 94 dB @ 4,250rpm

2006 Softail Standard 1,400 cc: [4] - 97 decibels at idle - 102 decibels at cruising speed - 111 decibels revved

So what it appears to me is that most motorcycles are `roughly` the same decibel output and that having a vehicle that outputs sound to make yourself known to other riders is beneficial. Should you be increasing the sound trying to be cool? No, I don't think so. In fact its annoying when the 5 Harleys are sitting outside my window on Friday night revving their engines to be "cool".

[0] - http://www.motorcycledaily.com/wp-content/uploads/2013/01/Qu... [1] - http://audiologyworld.net/img/rnhposter.pdf [2] - http://www.gencourt.state.nh.us/rsa/html/XXI/266/266-59-a.ht... [3] - http://www.autoevolution.com/news/honda-cbr250r-gets-street-... [4] - http://lic.abateflorida.com/Library/Library/cycle_news__nois...

Something vital is lost in motorcycling that my bike is so loud to make children cry and people really detest my presence everywhere in the name of saving my fat posterior.

All my bikes are quiet, even my Buell (relatively speaking). I have installed extra loud horns on them and I'm not afraid to use them. That's what horns are for: to say "I AM HERE", nothing more, nothing less. Save your engine for moving the bike forward, not making all of us bikers an anathema to civil society.

I agree with you, and I wasn't referring to ear piercing noise of modified bikes.

However, as I answered in a response to one of the other commenters, sound is an essential element in bringing awareness to yourself.

`All my bikes are quiet, even my Buell (relatively speaking)` - From what I've found in my short research is that most bikes are in similar decibel level output (within similar engine capacity) and that they are all fairly loud. Clearly there's something to that.

But I degrees, I never meant to say that you have to be extra loud to be an ass, I was merely pointing out that sound is an element of awareness, like your horn or the headlight.

If your vehicle can't be safe without annoying everyone within a quarter mile radius, it's too unsafe to be on the road.

Motorcycles should be illegal.

Like your spirit (I don't like the loud pipes either) but this is a little unfair. Motorcycling should be part of a more sustainable future. Especially when you consider how many OECD commuters commute in cars alone.

Except when you're in my blind spot, and your pipes are pointing backwards, I still won't hear you until you're in front of me and I've already seen you. Cars are incredibly quiet inside these days.

I understand that concept that "Your pipes are pointing backwards so the isn't helping with making people aware of you in the front". From my research, it seems that most motorcycle accidents occur with a driver crossing, to the left, in front of the motorcycle and that the decibels, horn, and lights didn't help the driver (of the motorcycle).

I also agree that you don't have to make your bike EXTRA loud to be noticed but there is a perceived safety in sound. And if majority of the sound is exiting towards the back, there might a point to be made that the decibel level should increase to "hopefully" make the person in front of you aware.

I hate loud motorcycles. I live in a quiet neighborhood in Alameda -- and I have small kids that sleep and take naps.

I dont like loud bikes driving by and waking them up :-/

I completely agree with you! I live in an area where on Friday and Saturday nights there are always groups of 5-6 bikers sitting outside and just revving their engines to "sound cool" and that's very annoying.

I didn't meant that you had to be extra loud or replicate the idiotic behavior of our neighborhood bikers. I was merely pointing out that loud sounders are safer. I had a more detailed reply above.

So right, too much noise = bad. Just enough sound = maybe good.

I agree. Please be active in the opposition to regulation requiring electric and hybrid vehicles to make a noise as well. We don't need more noise!

This is the silliest thing I've read all day. For starters, you're implying that drivers have less need to perceive small bikes...or other cars, apparently.

My reply comment wasn't very detailed and gained some criticism that I understand.

You make a valid point "... implying that drivers have less need to perceive small bikes...or other cars, apparently."

As I stated above, but will re-iterate because its a valid reply here, there seems to be a necessary of roughly 55db produced by a vehicle to ensure safety for pedestrians and other vehicles. [0]

My initial comment was incomplete: it's not okay to have an extra loud bike to be "cool", however, it is important to have a high enough decibel output that other people can hear you. Most bikes produce similar dB off the manufacturing plant for a reason and electronic vehicles are required to output more sound to be safe.

So no, i'm not implying that drivers have less need to perceive small bikes or other cards. I'm implying that all vehicles should produce the necessary decibel output to bring attention to themselves.

[0] - http://www.motorcycledaily.com/wp-content/uploads/2013/01/Qu...

When I last bought a VW (2008) I was concerned due to a clunking noise after first starting it up and driving a few miles; I called the VW service dept and they explained that there was a compressor which captured exhaust fumes and released them at a lower rate for the first few miles of each journey -- to help them meet EPA emissions standards. I was assured there was nothing to worry about.

While this is a different case, it struck me then that the emissions guidelines were being gamed by manufacturers. We're regularly reminded that barriers will be circumvented; what is the correct approach here? Like performance enhancing drugs in sports, too stiff of a fine and manufacturers are encouraged to find more subtle ways to beat the system, unenforced you leave a polluting industry to destroy our environment..

Perhaps the noise was poorly explained. What they describe is likely the Exhaust Gas Recirculation (EGR - https://en.wikipedia.org/wiki/Exhaust_gas_recirculation ). This isn't a system to game emissions, it's an active component in real emissions reduction.

Still I get your point: manufacturers are clearly going to optimize based on the measured metrics. How do we know those metrics are meaningful in the overall system.

ERG is just a pipe from the exhaust to the intake with a valve. Compressing and carefully releasing would be much more complicated and very much a not normal way of lowering exhaust temperature to reduce emissions.

Yes, on second thought, they probably confused secondary air injection ( https://en.wikipedia.org/wiki/Secondary_air_injection ) with EGR. It's very possible they were explaining secondary air, which doesn't compress exhaust gas, but does inject air into the exhaust gas to help burn more efficiently (not clear on the mechanism, I always thought it was a method to improve catalytic converter efficiency at startup).

I think the idea is that cold engines run rich, so injecting air into the exhaust gives the unburnt fuel a chance to combust in the exhaust pipe before getting to the catalyst. This is good for two reasons, first, it heats the catalyst up faster, which is good, and it also reduces the amount of unburned hydrocarbons.

The usual failure mode is that it sounds like a loud vacuum cleaner for ~90 seconds after cold start up.. not sure about clunking.

I owned an RX-8 which had an air pump (secondary air injection) which would come on for about 30 seconds on a cold start. When the pump started going out, it would make a metal grinding sound. Drove like that for a few years until it quit working, and didn't think anything of it until SMOG time - $600 for an air pump which works for ~30 seconds...

Exactly. Cheverolet Trailblazers are well known for this (I owned one and had to replace said air injector)

There is a more charitable interpretation.

When engines are colder, they burn less efficiently and produce "worse" emissions. A system that treats the warmup period differently than the steady-state of running a hot engine is potentially quite reasonable.

If you output those same emissions at a different time you haven't really done a lot for environment (except if they're cycled through catalyc converter or DPF filter which are usually less effective on cold start) - you're just gaming the system to get better eco ratings.

If you're recirculating exhaust back into the engine, you're able to attempt to reburn any fuel that wasn't burned on the first pass, while also exposing any emissions to higher temperatures on the second pass (higher temperatures = easier to break down when those emissions reach the emissions control system).

If you're recirculating exhaust back into the engine, you're able to attempt to reburn any fuel that wasn't burned on the first pass

That's not what EGRs do.

EGRs take exhaust, which is relatively inert (combustion having already taken place) and injects it into the combustion chamber to lower combustion temperatures. The reason you want low temperatures is that NOx (nitrogen oxides) form at high temperatures.

It appears I was mistaken!

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

"In internal combustion engines, exhaust gas recirculation (EGR) is a nitrogen oxide (NOx) emissions reduction technique used in petrol/gasoline and diesel engines. EGR works by recirculating a portion of an engine's exhaust gas back to the engine cylinders. This dilutes the O2 in the incoming air stream and provides gases inert to combustion to act as absorbents of combustion heat to reduce peak in-cylinder temperatures. NOx is produced in a narrow band of high cylinder temperatures and pressures.

In a gasoline engine, this inert exhaust displaces the amount of combustible matter in the cylinder. In a diesel engine, the exhaust gas replaces some of the excess oxygen in the pre-combustion mixture.[1] Because NOx forms primarily when a mixture of nitrogen and oxygen is subjected to high temperature, the lower combustion chamber temperatures caused by EGR reduces the amount of NOx the combustion generates (though at some loss of engine efficiency).[2] Gasses re-introduced from EGR systems will also contain near equilibrium concentrations of NOx and CO; the small fraction initially within the combustion chamber inhibits the total net production of these and other pollutants when sampled on a time average. Most modern engines now require exhaust gas recirculation to meet emissions standards."

That's traditional "external" egr. Many newer engines have an "internal" egr implementation, which just changes valve timing so that some amount of the combustion products are never evacuated from the cylinder in the first place. With cam phasing becoming commonplace, internal egr can be implemented without adding any additional parts, and is thus cheaper. So the "recirculation" part of egr is becoming more and more of a misnomer.

Personally I like internal egr because external egr systems have a tendency to clog in either the passageways or the solenoid controlled valve. EGR valve sticking is a common cause of cars stalling when slowing to a stop after cruising at medium-high speed. And even without the stalling egr is a common cause for check engine lights and smog test failures.

Car makers work hard to get emissions to be as stable as possible in all running states. For instance, cars have an air injection pump for the purpose of getting the CC to do its job sooner by increasing the temperature of the exhaust gas. It starts at a high RPM when you first cold start the engine.

Just because they used various tools to attempt to meet EPA standards doesn't mean any or all of those tools are illegal. What you're describing sounds like either a smog pump and/or a charcoal canister. Both are legal and widely used means of controlling emissions.

The emissions guidelines, in your case, likely weren't being "gamed" so much as the manufacturers were just coming up with innovative and cheap ways of reducing emissions. Start/stop features on some newer cars are a way of doing this but this doesn't sound like someone "gaming" the system. If you're not very automotive savvy then you may have misunderstood what the technician was explaining or the logic/purpose behind the system.

What VW did though is completely different with regards to the software although not necessarily illegal until we have all the facts. Based on what the article says, it sounds like they developed a special "emissions mode" which went outside the normal operating parameters with the ultimate goal of reducing the emissions. That in and of itself is not illegal since it passed emissions that way. The illegal part probably comes from when it switch back into normal mode where it never would have passed emissions that way.

We can assume for one reason or another that the "emissions mode" is unsustainable for regular driving (for one reason or another) which is why it was only used when absolutely necessary to pass emissions.

Huh? They disable "emissions mode" during all driving either because a) it was bad for sales - perhaps one is unable to reach race car speeds in < 10 seconds; or b) they purposely chose not to build a compliant emissions reduction system and covered it up by specifically making the cars pass only during examination.

There is no apology to be made for them. Manufacturers are required to build sustainable systems, it's not optional. Unless the initial facts turn out to be wrong, this is clearly inexcusable.

I see what part of what I said made you confused. The "unsustainable" part of what I meant was that it caused some sort of undesired effect be it affecting performance or causing accelerated wear on components that could be costly to replace under warranty or some other condition.

Either way, I agree, VW did this intentionally and there is no question as to their motives behind it. I'm sure they'll be severely punished for it.

That sounds more like a Diesel Particulate Filter and regenerator ( https://en.wikipedia.org/wiki/Diesel_particulate_filter#Rege... ), which is pretty much the opposite of gaming. To keep the black clouds of soot that have sullied the name of diesel from forming, most modern diesels are designed to capture the soot until the engine gets warm enough to burn off the soot, at which point said soot is passed back through the engine and incinerated. Or, there's some mix of heater and catalyst to burn off the soot in the regenerator itself. You're basically holding onto the waste until you can efficiently deal with it in an environmentally clean manner.

Some models of the Yamaha R1 have an intentional flat spot in their fuel maps which happens to be right at the spot where testing happens, too.

In a similar vein, the Aprilia RSV4 kind of cheats at noise testing: in neutral, there is a baffle which flips down to dampen the exhaust note a bit. :)

Almost all supersport bikes coming into the US utilize both of these strategies. :)

I own a Ducati 999 that was lemonlawed because it would stall at stop lights. Super common and happened because of how lean the bikes had to run to pass US emissions.

Google said "random stalls" were kind of endemic to Ducatis. :P Didn't know it was related to the fuel mapping! I had a 1098 which would occasionally (maybe once per month) randomly stall at 0% throttle when braking to a stop. I had a full Termignoni system + race ECU, though. California doesn't do emissions testing on bikes, surprisingly!

I had to put a Power Commander on my R1 because of the aforementioned lean spot. The engine would "chug" at about 37mph (5.2k rpm in 1st gear), which unfortunately happened to be a very common cruising speed. Google showed lots of complaints for other years as well.

I can assure you that the engine on the R1 doesn't chug along at sport bike speeds though ;)

Yes, certainly not. :) https://www.youtube.com/watch?v=HSsP-itS8qI

I commuted to work in San Francisco on literbikes, but on weekends I would head south and open them up.

>the emissions guidelines were being gamed by manufacturers

Not necessarily gamed - the measuring just needs to factor it in. e.g. (Rather anecdotal) I know there are cars programmed to capture emissions in a filter & then automatically clear that when the car reaches high speeds (freeway). Makes sense in a way...you'd rather have the car expelling the crap on the freeway than city center. That mechanism is apparently a big problem here - I live in a place where the cars never reach freeway speed so they keep clogging (so I'm told).

I believe you mean the EVAP system where excess gasoline vapor is condensed in a charcoal canister, this canister is purged generally in a freeway with a high engine vacuum going on. the highest vacuum is achieved when you engine brake

Just like sports, even stiffer fine and ruined reputation on being caught cheating.

How about a temporarily sales ban? Not a fine, just telling them "No new VW are going to be accepted into traffic for the next 12 months."

Cruel, but I bet it's effective.

That's not a good idea. If you ban a large auto maker, you might create a shortage of vehicles, driving prices up for consumers. Also, the affected auto maker would surely try to mitigate the damage by cutting costs, closing down factories, firing workers. Dealerships would have to close as well.

In other words, you would cause a lot of collateral damage instead of taking it directly from shareholders via a large fine.

Every time you arrest a parent for committing a crime, you destroy a part of their child's life. Using collateral as justification to avoid punishment results in a system where entities can become to important to punish at all.

We're talking about how to punish not if. If you can punish a parent in a way that does not affect the child, surely that has to be preferred way of doing it.

> That's not a good idea.

Agreed. If auto makers face the risk of significant fines and/or penalties they will have to spend more on compliance. The worst thing for innovation is to create an environment where engineers are effectively being managed by lawyers and accountants.

You just have to make sure that the fines are not tax deductible...

http://www.nytimes.com/2015/02/04/business/when-a-company-is...

Just like sports, even stiffer fine and ruined reputation on being caught cheating.

Just like sports, companies will still cheat. When punishment doesn't work, the answer is rarely more punishment.

Caveat: if the punishment does not exceed the profit from cheating, it's not a punishment it's a cost of doing business.

See: Johnson & Johnson w/ their drug risperdal: $2 billion in penalties... $30 billion in sales.

http://www.nytimes.com/2015/09/17/opinion/nicholas-kristof-w...

That's a poor comparison. Not all of the $30B in sales were fraudulent.

Actually, in order to come up with the $2B penalty, the gov't determines the number of fraudulent Rxs, then hits the manufacturer with triple damages (this is typical).

J&J would have been better off financially had they never committed the fraud they did.

You forgot to multiply the cost if caught by the chance to be caught.

Of course it is. Companies usually behave rationally with regards to regulation and punishment. If the cost of breaking the rules is lower than the cost of following them, companies will break the rules. If you increase the cost of breaking the rules such that the situation is reversed, then companies will follow the rules.

The tech industry has been guilty of this in the past as well. Example: Nvidia and AMD drivers detected if you were running certain benchmarks or games, and would cheat by downgrading various rendering settings that had minimal effect visually.

Often this was a simple check to get the process's executable name; simply renaming the .exe would get you significantly different perf results.

Yeah that was how I first learned about this with ATI cheating on Quake III in 2001, as discovered by HardOCP. Here's a link to an article about it (original article doesn't seem to be around anymore):

http://techreport.com/review/3089/how-ati-drivers-optimize-q...

Wayback has it http://web.archive.org/web/20020223144052/http://www.hardocp... along with a link to a German site that points out some of the quality loss in the ATI driver. http://alt.3dcenter.org/artikel/2001/10-24_a.php

PhysX's "CPU-only" codepath was intentionally built at a lower optimization level than their "with NVidia GPU" codepath for a while, leading to disproportionately lower performance than simply the lack of GPU kernels would create:

http://semiaccurate.com/2010/07/07/nvidia-purposefully-hobbl...

Mobile device manufacturers were doing it 2 years ago: http://www.anandtech.com/show/7384/state-of-cheating-in-andr...

In this case defeat devices are specifically outlawed though, and VW have admitted creating them, so they are actually going to get punished, may put some people off.

MSFT was accused of gaming javascript benchmarks with IE9.

This was a longstanding practice to make diesel engines more efficient while passing the official emissions test. There was a similar enforcement against Caterpillar and other truck makers in 1998:

http://www2.epa.gov/enforcement/caterpillar-inc-diesel-engin...

And against Ford in 1998:

http://www2.epa.gov/sites/production/files/2014-06/documents...

along with a clear ruling that defeat devices were illegal.

I wonder if the primary control code should be open source. That's without respect to right to modify or reuse the code, just whether it should be mandatory that it be published. And it'd exclude things like user space apps and UI/UX stuff. But any of the code that controls or affects steering, acceleration, ignition, braking, and lighting, really seems like it's not OK to just consider it a black box, subject to DMCA, copyright, and trade secrets laws and thus risky (or flat out illegal) to reverse engineer and understand how it works. And right now a significant part of an automobile is a black box, and you do not fully own it.

It was supposed to be that patenting required you to fully disclose all the details of an invention. The idea was to not only make clear what you had a monopoly for, but so that others could learn from your invention and make something different that was better.

If it's just published you can't be sure it's the same code that runs in the car. You should be able to compile the code and inject your locally compiled firmware to the ECU if you want to verify that the car engine is doing what it's supposed to be doing.

If you can do it, chances are an attacker can, too. I prefer hardware that won't load unsigned software in such places.

Then use reproducible builds. You inspect the source code, and if it does what it's supposed to, compile it reproducibly, and compare the binary with the official one which has a signature. If they match, then everything is kosher.

Difficult unless the compile tools are free and open source, and the manufacturer documents the exact version of what compiler they're using, and what feature flags are enabled, etc. Otherwise the exact same source code can produce many different binaries, each of which hash differently.

I don't know the solution, but the black box approach we have right now I do not like. As a pilot this is a bit weird for me to say because those systems are also completely proprietary as well, full on black box - and not the d/v recording type. And I'm even wondering if that's overdue for a change as well.

Perhaps we need an approach whereby abuse of patented or otherwise protected technology should be punished in part by termination of the monopoly protection, and its immediate free availability to competitors. I imagine it would create some serious economic hurt for VW if these allegations are proven to be true and their ECS were made public domain - far more than a fine, or at least any fine that's likely to be politically acceptable.

This is actively malicious. There had better be criminal charges filed.

Harm to the consumer via fraud, and to the environment, yes. But corporations almost never see criminal charges. Fuck the environment for 6 years? Pay this fine. Sell some pot? Fuck you asshole go to jail for 3 years.

Well, for the BP spill at least, people were charged with negligence, but the trial is still under way.

Criminal charges seem like a good plan. But its not sufficient.

Volswagen has over a period of time pumped a lot of this gas into the atmosphere and the fine ought to reflect the actual clean up damage. What is it going to cost to remove it and put the environment back to the way it should have been if they hadn't broken the law. That is what these sorts of fines ought to be based on.

No fines should be punitive so they dont do it again. It is not about clean up with particulates, it is about their medical effects, which you cant be specific about.

Fines that would really be enough to clean up the environment would probably obliterate the company.

Capital punishment is a deterrent against rational actors.

My point was that a punitive fine was not necessary (as a deterrent or otherwise), but a fine that realistically represented the clean-up costs would already be enough. (Destroying such companies is fine with me.)

Agreed. This somewhat mollifying quote gives hope:

>"Using a defeat device in cars to evade clean air standards is illegal and a threat to public health," said Cynthia Giles, the E.P.A.’s assistant administrator for the Office of Enforcement and Compliance.

Also, this:

> California, the E.P.A. and the Justice Department are working together on an investigation of the allegations.

Though I hold similar pessimism to cmurf. If we can't even bring criminal charges against people that destroy the economy, I don't have high hopes for anyone being charged for destroying the environment (given the difference in priority they're normally given).

I expect at the least they are going to get hit with the maximum administrative fine, 37,500 per vehicle.

That maths out to somewhere in the ballpark of eighteen billion dollars.

In 2012, the EPA issued $252 million in fines, total. Kia & Hyundai were fined $300 million total in 2014.

I don't see them hitting VW with eighteen billion.

e: Eh, that might not be so unreasonable after all. In 2012, VW made about $28 billion in profit on $250 billion of sales.

If the cars can't pass smog testing without cheating, then they also need to be forced to buy each and every one of them back. $18 billion sounds reasonable to me. This needs to hurt.

I'm sure the patch will allow them to pass the tests. It'll just reduce driving performance as well.

Likely true; the problem with that is the owner ends up with a car with lower specs than they agreed to when they bought it. The owner of the car needs to be compensated.

it would be an insane judgement seeing how little GM paid out for a defect they willfully hid that led to actual recorded deaths.

Which is $18b, so not insignificant. There may also be action in Europe.

$18b would mean they’d still make profit out of this.

And it’s not like only VW did this, every carmaker currently does this. It’s not going to change soon either.

No, they dont make that much profit per car. It could usefully be a few times higher (punitive damages and criminal charges could do that though).

I would like to see VW prohibited from selling any vehicles in the US for a small number of quarters. That would be real punishment.

All those dealers and salespeople? There are a certain amount of cars that need to be sold. If VW is out of the mix, the demand is still there. The dealers can sell something else. I'm certain that Toyota and Ford can quickly come to some arrangement with those dealers, including any state legal issues.

Do you have even the faintest clue of how car sales work? Or of how regulatory enforcement works? The EPA can't ban VW sales entirely.

I believe the AG's office is involved.

Anyone know if this "defeat device" would be a violation of the various Euro* regulations in Europe? It's been widely reported [1] that diesel cars emit more NOx than the specification allows, but so far no consequences. More than 50% of all cars sold in the EU are diesel fueled, and many countries are breaching the EU air quality limits. This causes an estimated 400,000 premature deaths each year [2]

[1] http://www.theicct.org/news/press-release-new-icct-study-sho... [2] http://www.euractiv.com/sections/health-consumers/eu-sends-b...

Apparently made illegal at least by 2007 regualtions (maybe adopted later, not expert in EU law)

http://eur-lex.europa.eu/legal-content/EN/ALL/?uri=CELEX:320...

What actually incentivizes a customer to take their car to a shop to get this fixed though? It's not causing harm to individual car owners, and it's not like they can get collections agencies to sieze them.

Seems like they're going to have trouble getting compliance here...

In California CARB can order all of these cars to be inspected and if they aren't, the DMV won't renew the registrations.

OK, didn't know if that was something that was done or not (see my response below).

I'm in CA, just had to renew my registration and get my car smogged a couple months ago (its a 2013). It passed, of course.

I've got a 2013 Golf TDI and was thinking about this as well. If it isn't a safety related recall, I'm not sure how you get people to comply aside from some other action related to, say, their next years registration (i.e. if your VIN doesn't have this fix applied, you can't reregister your car). Not sure if that's a thing, though.

I'd personally plan to take my car in "real soon now" to get the fix.

I wonder how you car will drive after the "fix." I suspect you won't be very happy with it.

That seems like a big liability for them... I'm sure some enterprising lawyers are already plotting a class action. Consumer buys car after test drive and loves performance, but it turns out performance was a lie. They may have to do a buy back.

The vast majority of consumers won't notice. Some will be upset, because They bought the car for the specs on paper just to feel cool. Similar to how the fanboi segment of consumers by video cards and computers and stereo equipment.

Yup, quite possibly thinking about bailing on the car sooner rather than later. I have no idea if this kind of thing significantly affects resale value, but I'm not sure I want to wait around to find out.

There will be aftermarket kits to remove the controls. Just like there are now for all diesel engines.

I have a 2014 Jetta TDI, and the thing I'm wondering is what the cost of the fix will be - not to me in dollars, since it would be covered by the recall - but rather, will this affect my fuel economy or performance in some other way? I'm just trying to figure out why they don't have the cars running in this lower-emissions mode all the time.

The performance and/or lifetime of the engine will likely be compromised to meet the standards.

Epa can potentially require VW offer an incentive to owners.

I don't understand how California detected this. The one time I had to take my 2010 Jetta turbo diesel for a smog check, the inspection was ENTIRELY visual. The inspector did not insert a probe into the tailpipe. He did plug the test rig into the OBD-II port under the steering column and turn the key to the accessory position (presumably to electronically capture the VIN), but that's it. I even asked him at the end if it was an entirely visual inspection and he said, puzzled, yes but that's what the computer told him to do.

California didn't originally detect this. According to the letter from the EPA [1], it was a team at West Virginia University doing a study that discovered that TDI's in normal usage were found to have significantly higher emissions. California's CARB then dug into it to confirm their results and work with the EPA to figure out what was happening.

[1] http://www3.epa.gov/otaq/cert/documents/vw-nov-caa-09-18-15....

California has a new emissions check system for vehicles 2000 and newer. No tail pipe probe or dyno regardless of the vehicle. Just visual inspection and plug in the OBDII reader. The OBDII reader checks the sensors for proper functioning and checks the ECU to make sure it hasn't been tampered with. Smog tech only really has to do the visual chek. Diesel trucks also get a snap test (step on the accelerator quickly and diesel smoke must dissipate within several seconds).

Vehicles 1999 and older still get the tail pipe probe and dyno portion as well as the OBDII (if they have OBDII).

[0] http://www.bar.ca.gov/Consumer/New_Smog_Check_Test.html

[1] http://www.smogtips.com/new-smog-law-ab-2289.cfm

Yea, they do get the tailpipe emmissions test. Most shops use just one reference book for the visual inspections.

The shop owner usually, almost always, buys the $100.00 Motor Emissions manual.

The cheap shop owner should buy OnDemand5(Mitchell Manual Smog Emission) subscription/publication. They don't because they are cheap, and CARB only requires one reference manual for customers, and technicians.

The problem is Motor Publications is filled with errors that send thousands of customers home with failed Smog Checks.

I have found so many errors in Motor publications, it's comical. I have numerous emails, where I am literally arguing with, I believe, a secretary, over errors in their manuals. I have one email when the person I'm communicating with doesn't know the difference between Naturally Aspirated(Carburated) v.s. Fuel Injection. It's beyond frustrating, and I hope they get sued!

If you live in California, and your car doesn't pass the visual smog check, it's probally not you--it's errors in Motor Publications?

I have found errors on older Toyotas, and Volvos. Just the two autos I happen to own? What are the odds of that? They failed my vechicles, until I went to a shop that had Mitchell manuals, and I presented the Techs with a Printout of the Correct Emmission information.

(Motor Emission manuals--If you are listening; please clean up your books? It's beyond frustrating to fail a smog test because of your sloppy research! I will show the world my frustrating emails--if necessary?)

(CARB--if you are listening, require every shop to have access to two (2) separate Emmission manuals from different entities?)

And no--1988 Toyota trucks, 2wd, naturally aspirated(Carburator) do Not require a MIL light.

I literally tore apart the dash looking for the missing light. There was no light. There was no plug for a light. I literally thought I was going crazy. I was so frustrated, I forget to set the emergency brake and the truck rolled out of the garage. I'm lucky I wasen't on a hill and no one got hurt. I did need to buy a $200 bumper though. I'm still thinking about taking the shop to small claims court, but don't think it's worth the effort?

So, if you live in California, and fail Smog--ask questions. Don't just bring it to another Smog testing station praying it will magically Pass.

Naturally aspirated means the engine breathes air at regular atmospheric pressure - ie is not turbo or supercharged.

The type of fuel delivery has nothing to do with whether it is naturally aspirated or not. A NA engine can have fuel injection and a supercharged engine can have a carburettor.

California's CARB is the gestapo of car emissions. If anyone has the motivation, expertise, and evil disposition to detect this, it's the state of California. I've never lived in a state like this. Bi-yearly tailpipe sniffs and visual inspections. Even if your car's emissions are cleaner than their ridiculous requirements, they'll still fail your car if it has parts on it that are not stock. I ended up having to sell my car to someone out of state because I'm not going to swap in and out my carefully customized (and clean) intake and exhaust system every two years just so they can shine a flashlight in and see stock parts.

I'm not shocked that California is a part of this.

Very happy to hear CARB is working!

ryandrake's point was probably that his car is cleaner than necessary, but failing because of stupid, arbitrary rules imposed by CARB for no reason other than bureaucratic intransigence.

It's similar to the feeling I got when I waited in line for an hour to have my Miata smog-checked at a busy urban test station, and then choked on diesel exhaust from a dump truck when I made the mistake of driving away with the top down. CARB is the poster child for "Doing something just to say we're doing something."

Bureaucratic intransigence is a really upbeat way of putting it. How about corruption through regulatory capture? Big auto needs to sell their over-priced parts. Why let people buy a $150 after-market part, when you can just call up CARB and have them make people buy equivalent $1,100 factory or OEM parts?

Wow, judging by the down-votes, people don't seem to realize how expensive CARB deliberately makes it to own an older car, regardless of how clean its emissions are.

EDIT: To further demonstrate that it's all about money and not about smog, you can get a "Repair Cost Waiver"[1] if you've spent at least $650 (on CA-approved parts, of course), and they'll approve your polluting car, no problem. All they really want is that $650 in the pockets of their car industry buddies and $XYZ worth of repeated tests in the pockets of the test stations.

1: http://www.bar.ca.gov/Consumer/Referee/Referee_Centers.html

Should we fault CARB though? They require more stringent emissions regulations due to California's geography, and both the federal government and automakers are rarely amiable to their needs.

With cars going electric, its a bit of a moot point, except for the part where pressure needs to be applied on heavy trucks now (hopefully pushing them to natural gas from diesel until energy storage catches up).

The reason California is so strict is because LA used to be synonymous with lethal levels of smog.

Used to be.

I find it difficult to understand how failing a car with clean emissions levels, thereby forcing it out of the state, contributes to cleaning up smog.

The point is that California had an exceptionally bad problem, which is why the response was so stringent.

If it's forced out of the state the smog isn't California's problem. California's emission regulations aren't meant to clean up other states.

You are willfully ignoring the vastly more significant average/mode case to focus on the effect of edge cases.

You're not wrong, but tailpipe tests were quite common in many states. In CA, like other states, they're being phased out.

Nevermind the fact that California has no safety inspection, unlike many Eastern states in particular. It's fairly silly to call CA the "gestapo" although they are more aggressive about emissions than anyone else, as you mentioned.

I am a driving and motorcyclist enthusiast, and have modified cars and regularly participate in track days. Somehow I've never had a problem with CA emissions, nor have most of my like-minded friends. It's just not that bad, and it does a lot of good.

More interestingly: Did the German TÜV not detect this? They have even stricter rules.

And completely different emission and OBD requirements. I don't know how much hardware they share, but they definitely use different software.

Yes, but the TÜV should have noticed such faking, too.

Europe doesn't have stupid emissions rules biased against diesels.

On the contrary, California emission standards held diesel passenger vehicles to the same standards as gasoline/petrol passenger vehicles. Whereas in Europe, diesel emission standards were much more lax compared to their petrol counterparts.

I suppose one could say that being more concerned with the more directly harmful and smog forming emissions(NOx and PM) over CO2 would constitute 'bias' against diesels. But, conversely, there's a rather compelling argument that the more lax European diesel regulations spurred diesel adoption and protect their domestic manufacturers.

https://www3.nd.edu/~jthurk/Papers/MMT.pdf

http://www.autocar.co.uk/car-news/industry/diesel-dominance-... http://www.theguardian.com/environment/2015/sep/14/nine-out-... http://www.bbc.com/news/business-34257424 http://www.hybridcars.com/will-america-avoid-europes-clean-d...

As I understand it you don't get the full tailpipe test unless your vehicle is too old to support the ODBII parameters they look for.

Colorado does drive-by testing, on freeway on ramps, and other locations with high traffic. If your car fails, or hasn't been monitored by that system recently, you get a notice for the more complete inspection.

The drive-by inspection doesn't require the co-operation of the driver or car, so is harder to game by detecting you are being tested.

As a TDI owner and Colorado resident, I can tell you that the drive-by testing does not apply to diesel vehicles, nor do you take diesels to the emissions testing centers. You have to take your diesel to a state-licensed diesel mechanic for testing.

I'm not sure about CA specifically, but some states only do full inspections on newer vehicles every other year. e.g. In a particular year the odd year models get the full inspection and the even year models just get a quick inspection and then vice versa in the next year.

ca doesn't even do smog checks on newer vehicles (i believe 5 years, not sure though).

i haven't done one in over a decade because i lease my cars as a business expense.

also, ca has 'test-only' smog centers which do not perform repairs - if you fail there, you have to get your car tested and repaired somewhere else.

needless to say ca highly incentivizes you to buy new clean cars. it quickly becomes very expensive to fix an old, polluting car under this regime. imagine how fast the bills can pile up if the first or second fix doesn't work. with new econo class cars available for ~$100/month on finance or lease, it doesn't really make sense to keep an old clunker around unless it's a collectible.

The 6-year exemption for new vehicles does not apply to diesels. Diesel cars are required to get smogged every two years.

incentivizes you to buy and old clunker ( pre emissions ) and keep it running forever... or a truck...

Interesting to note is that 1997-older diesel trucks/cars did not and still do not require any smog testing at all.

"Currently, smog inspections are required for all vehicles except diesel powered vehicles 1997 year model and older or with a Gross Vehicle Weight (GVWR) of more than 14,000 lbs, electric, natural gas powered vehicles over 14,000 lbs, motorcycles, trailers, or gasoline powered vehicles 1975 and older."[0]

[0] https://www.dmv.ca.gov/portal/dmv/detail/vr/smogfaq

Georgia doesn't even test diesel cars, others only if over three or four years old.

And only in the metro ATL area. The rest of the state doesn't do emissions testing.

Wisconsin doesn't test any cars older than 1996, because they would need dynos for it. Much cheaper for them to just only test ODBII cars.

Here in Oregon my '03 TDI gets a quick ODB-II and a visual inspection (i.e. is the Check Engine light off, is the exhaust not full of smoke, etc.). Of course, I'd expect CA to be the most stringent state.

OBD-II tracks emissions data as well. http://ctemissions.com/test-procedures/what-is-obd

i had to do a smog test for my 1 year anniversary of owning a new car. it is a 2015 model from a different manufacturer, acura, in my case, and it was visual inspection only. i asked the guy, because i'm used to having actual measurements performed each time. if a car is new enough then they only do a visual inspection. i don't know the details of how "new enough" is determined, though.

i've seen this also when having fire extinguishers inspected; ones that are only a year or so old are also only visually inspected only.

In NY the first year is safety only; but year 2 and on are emissions + safety (depending on the county you register your car in, some never need emissions). The safety inspection only costs $10; IIRC the emissions + safety is $37.

Newer cars tend to have more relaxed testing requirements... in many places cars don't even require smog checks for the first five years.

They even made commercials to prove how clean the cars are. https://www.youtube.com/watch?v=WNS2nvkjARk I own a TDI VW and fell into believing that the ride is indeed clean. Now it looks like their software hacks are even more refined than the combustion tech.

Of course this is a major part of their marketing. If the EPA makes them reflash my ECU and as a result my fuel consumption goes up noticeably I'm going to expect them to buy the car back. They sold them on false claims.

This also pretty much guarantees that I will never buy another VW.

> If the EPA makes them reflash my ECU and as a result my fuel consumption goes up noticeably I'm going to expect them to buy the car back.

Good luck with that.

If they face a choice between having to pay an EPA fine equal to the original sale price of the car, or buying them back at depreciated values, I can imagine which one they'd prefer.

Well, at least we've finally found out where the "Partial" in PZEV comes from.

If the ultimate and only resolution to this is that VW has to pay a fine, or other outgoing monies due to the recalls, in an amount that is not significant to the operation of their business, nobody should be surprised that this happened, nor question whether it can happen again.

I don't quite understand how the car can known it's getting an emissions test. I thought you just pulled up and revved the engine. Or do they plug the computer in nowadays?

In Wisconsin you drive to a specific testing facility and, when it's your turn, onto a testbed with rollers under the wheels. They spin endlessly without the car going anywhere until the test is over, when the rollers are locked so the car can drive off of them. During the test they attach a hose to the exhaust.

Engine revs in drive with no corresponding accelerometer output or change in GPS position should be a pretty solid indicator of "I'm in an emissions testing facility right now." Doubly so if there's an OBD-II reader attached.

Where I live, they just plug a thing into the OBD-II port and go with whatever the car tells them. (I drive a '96 Honda Civic.) They don't test the exhaust directly.

Uhm, the emissions test you do every year (I guess in some US states is the same as in EU where you have to measure that yearly?) is different than emissions test that's done on each new model of car before allowing them to be sold and then entered into official specifications.

Remember there are restrictions like EURO5 etc. that impose higher taxes, city environmental taxes and even outright ban cars from being sold on the basis of these specifications.

Yeah, plug in OBD-II reader, get result back.

Some emissions places also do the tube-up-tailpipe check, but the car could adjust exhaust output if it has an active OBD-II connection.

Tests have fixed usage patterns. You detect those while total mileage is low, and you assume it's a test. After X thousand miles, you stop detecting. Job done.

eh, not really. In atlanta, it was once a year, around your birthday. Where I am, outside savannah, it's never. So my car would have no idea if it's being tested or not

That's not the test we're talking about here. This is the official test to get a certificate of compliance and sell a vehicle in the US.

I can second this, once a year in the Metro Atlanta area. Where I went to college they didnt do emission. Many classmates switched their registration so they didnt have to pay yearly emissions.

First of all, that's not how it works. Secondly, being in neutral and revving the engine seems like a pretty obvious red flag for "hey, maybe I'm being tested for emissions" in the scenario you mention.

The NYT is being disingenuous in presenting this as VW detecting the EPA test cycle and doing something to cheat. I think they've confused open loop and closed loop engine operation.

VW's 4-cylinder engines which failed the EPA tests are all turbo engines. Turbo engines run hotter than naturally aspirated engines and so, they produce more NOx, since it's one of the high temperature byproducts. In fact, EGR (exhaust gas recirculation) systems inject exhaust into the intake to lower combustion temperatures and NOx output. All engines do this, turbo or not.

Whatever NOx makes it out of the engine can be catalyzed in the catalytic converter if there is enough O2 for the catalyst. All the engine's intake air and temperature sensors, and pre-cat and post-cat O2 sensors exist to balance the gases coming into the catalytic converter to optimize the catalytic reaction.

When the engine is using all of these sensors to fine tune the amount of air, fuel and EGR to maximize catalyst efficiency, it's running in something called "closed loop" mode. The sensors and inputs work in a closed loop to balance the system. This is what happens when you're cruising on the highway or driving around at relatively fixed throttle.

Give the car a lot of gas, and something else happens. All of a sudden, a LOT more air enters the engine. The sensors can't change reading fast enough (mass air flow sensor, for example) or their readings are now out of the sensor's useful sensitivity band (for O2). The engine code knows that sensor capabilities have been exceeded, and it goes into "open loop" mode. In this mode, rather than using sensor readings, the engine reverts to pre-programmed maps, usually indexed by RPM, throttle position, engine load, and boost level. These maps are designed for relatively clean combustion and engine safety, but they will not generate optimal catalyst input gases, and so, the catalyst efficiency drops off tremendously. Furthermore, performance cars will often burn rich (too much gas for amount of air) in order to prevent engine destroying pre-detonation.

Most of the EPA test cycle will happen in closed loop mode. Most high throttle acceleration or "fun" driving will happen in open loop mode. There's the source of the discrepancy between EPA test cycle and real world driving. VW got caught, but every car does this, as we don't have wide-band, high-rate sensors yet at a price point even close to where it would need to be for consumer cars.

(Next time you see someone racing off from a stop light, in pretty much any car, watch the exhaust pipe, you'll see dark smoke. I guarantee you they're running rich in open loop mode).

> The NYT is being disingenuous in presenting this as VW detecting the EPA test cycle and doing something to cheat. I think they've confused open loop and closed loop engine operation.

The NYT is not at all confused. The cars in question have software in the ECM specifically meant to detect testing. The software looks at various inputs, including steering wheel position, speed, duration of operation, and barometric pressure. The values it looks for precisely track the parameters of the federal test procedure use for EPA certification.

When the software detected this, it switched to a mode that VW actually called "dyno calibration". At all other times, it used a mode VW called "road calibration".

Source: http://www3.epa.gov/otaq/cert/documents/vw-nov-caa-09-18-15....

Wow. From the source:

"VW continued to assert to CARB and the EPA that the increased emissions from these vehicles could be attributed to various technical issues and unexpected in-use conditions. VW issued a voluntary recall in December 2014 to address the issue. CARB, in coordination with the EPA, conducted follow up testing of these vehicles [...] to confirm the efficacy of the recall. When the testing showed only a limited benefit to the recall, CARB broadened the testing to pinpoint the exact technical nature of the vehicles' poor performance"

I remember spending half a day at the dealership to deal with this recall, which was required to get my CA registration renewed. Can't believe it was just a BS update that didn't really fix anything. Shocking and extremely disappointing behavior from VW

Wow, jumped the gun on giving VW the benefit of the doubt before reading the actual complaint. Thanks for the link.

So the VW ECUs essentially have a separate map for when they detect the EPA testing.

The cars in this recall are all diesel.

How would this benefit Volkswagen? The article says the device activates emissions control systems when it detects the car is being tested... so it's not as if they're saving money by emulating a system they're not manufacturing.

Reducing emissions usually comes at the expense of worse power or fuel economy.

By only reducing emissions during testing, they can demonstrate better engine power and fuel efficiency while also claiming to meet the emissions requirements.

Don't the emissions systems affect performance? I have a 2012 Golf TDI and really like it, one of the advantages over one of the hybrids is that I get pretty great fuel economy for the kind of driving I do, but it doesn't feel like I'm driving a weed eater.

This really pisses me off. It doesn't surprise me I guess but it never occurred to me that a car manufacturer could get away with something like this for, apparently, 6 model years.

Yes, a common strategy is to inject more EGR (exhaust gas) into the fresh intake air. As EGR is already burned, it is inert, so it negatively affects the combustion, lowering the temperature. With lower combustion temperature you get less NOx... but less power for each amount of fuel, so your efficiency/performance is lower, and also non burned products (CO, C) increase. You then trap those with the Diesel Particulate Filter... it's a very complicated game with very small margins, that's the reason of so much cheating.

Since this EPA action is about NOx, it seems VW will need to lower combustion temperature, leading to higher soot and therefore more particulate filter cycles. That will also increase fuel consumption.

EGR on a diesel also terribly gunks up the intake manifold downstream of the EGR introduction point. For that reason (and the loss of power and efficiency), it's often defeated.

To pass the test with lower emissions. This means lower taxes in some countries, and good PR in every one. Edit: to expand this, consider that in order to reduce emissions like NOx, you often must sacrifice performance, so the customer will complain. When standarised tests were introduced, a crude counter strategy was to trade performance for lower emissions for X amount of minutes (the duration of the test). So customers noticed "hey, my engine works much better after 15 min!" :P

It's possible that the engine power / automatic transmission shift profile / maximum RPM / etc. is being manipulated to output less emissions at the cost of acceleration, top speed, ride comfort, etc.

The most fuel-efficient / eco ride is rather "boring" by the standards of car enthusiasts and reviewers and I guess the cars wouldn't be configured for that mode by default.

(Just a guess though.)

Or, you just pay more for a car that has those performance characteristics both in taxes/registration fees and in fuel.

That said, cheating like this is really bad on a very large scale, and should probably face fines and maybe criminal charges, unfortunately discovery is likely to fall outside U.S. juristiction so it would be hard to prove enough to bring criminal charges against the appropriate managers and executives.

It probably ruins performance -- that's the usual tradeoff with emissions systems.

Help sell cars. I have a TDI sportswagon and this will be the second performance hit. A couple years after purchase there was a recall to reduce the power a bit via software "upgrade". Seems fuel filters were occasionally failing at high acceleration or with people that used high acceleration a lot. The power drop was minor but noticeable. I suspected then and even more after this indecent that VW knew about the problem but wanted max power to help sell cars during the test drives.

The VW/Audi 2.0t EA888 engine is capable of producing 100 more hp with a software update. These turbo/diesel/supercharged engines have a million parameters that they can modify to change driveability. I cant believe that they would do something like that to pass emissions testing though, that seems incredibly foolish. I would imagine that with diesel engines, economy and certain emissions are at odds with each other.

They are, in particular the higher the in cylinder combustion temperature the higher the efficiency, and the higher the NOx production. This caused heavy diesels to get less efficient for a few years when NOx stringencies were tightened. Then they were tightened far enough to require catalytic reduction, and manufactures went back to high efficiency high NOx combustion and removed the NOx downstream.

My guess is it uses some part that would need to be replaced (or replaced more often) if left in full use all the time ?

That could be—especially since automakers are required by the EPA to include emissions warranties of 8 years/80,000 miles.

I guess this is the automaker equivalent to a CPU detecting its running benchmark code.

> The software was designed to conceal the cars’ emissions of the pollutant nitrogen oxide, which contributes to the creation of ozone and smog.

Is this an error in the article or is creating ozone a bad thing?

> is creating ozone a bad thing?

At the ground level, yes.

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

"Ozone is a powerful oxidant (far more so than dioxygen) and has many industrial and consumer applications related to oxidation. This same high oxidising potential, however, causes ozone to damage mucous and respiratory tissues in animals, and also tissues in plants, above concentrations of about 100 ppb. This makes ozone a potent respiratory hazard and pollutant near ground level."

Ozone is a respiratory irritant when it's at low altitude, so pumping out ozone isn't a great way to help the ozone hole:

https://en.wikipedia.org/wiki/Ozone#Ozone_air_pollution

It is! https://en.wikipedia.org/wiki/Ozone#Health_effects

Ozone production at ground level won't reach the upper atmosphere. It is too reactive. So all it's doing is burning people.

> Ozone production at ground level won't reach the upper atmosphere. It is too reactive.

And, also, heavier than air.

Nitrogen oxide is the new smoking: http://www.telegraph.co.uk/motoring/news/10862975/Emission-t...

(Article is from last year - it's just something I found on google just now - but representative.)

I have an Audi TDI, but luckily not one of the affected models. I would be furious, especially if post-update performance was noticeably different. If so they are going to be up to their eye balls in class action suits demanding buybacks and the like (not unprecedented, Dodge just did a buyback).

I drive a mazda, and my car definitely runs rich (visible buildup on tailpipe, the exhaust smells like gasoline when the engine is cold), but it just passed its first emissions test (at 6 years old) showing no detectable gasoline. I wonder if there's some gaming going on there too.

From the factory all vehicles run rich - it helps immensely with longevity as the extra fuel cools and lubricates the engine.

Vehicles run rich when cold and at full throttle. No engine would pass emissions checks if they didn't run stoichiometric mixture during idle and low-throttle operation.

> stoichiometric mixture during idle and low-throttle operation.

I don't believe I've ever seen a factory tuned vehicle run less than 16:1, most are 17 or 18 : 1

That was an accurate statement like 40 years ago.

Many of the comments in this thread follow a similar pattern. They are all essentially derivatives of Goodhart's law. [0]

>> Once a metric becomes a target, it ceases to be a good measure.

It's quite useful to have a name - a label - for this sort of behaviour. In this case, the means by which VW optimised for this metric was crafty computer software! A metric is only as good as the cost of the unintended consequences that come with imposing it on people, in particular large corporations.

I am not surprised that this is done by many corporations as mentioned elsewhere in the comments (ATI/Nvidia etc). This is because regulations and benchmarks that can be gamed, such as engine efficiency, will be gamed because the incentives of corporations work out that way:

(I hope this isn't perceived as an anti-corporate rant because it's not. Just a way of framing facts).

For regulations imposed on large corporations in general (think EPA admissions standards):

1) Regulations are mandated by a government to effect some purpose (stop destroying Earth)

2) Executives - the people running the corporations - are bound by legal responsibility - their fiduciary duty [1] - to create value for their shareholders (selling lots of cars, and profitably).

3) Regulations can obstruct the activities of a corporation by making them do things they don't want to (making efficient cars is hard and costly and affects the bottom line - its cheaper to manufacture a car with less platinum [2] in it)

4) Companies are incentivized to minimize the time and money spent on adhering to regulations

5) Once a regulation is standardised and codified, it becomes possible to follow the letter of it, while ignoring the spirit of it. All a corporation has to do is meet the metrics in the specification to abide by the regulation - in other words - game the system.

Given these incentives, we have seen that corporations prefer to abide by regulations in the maximally efficient way: gaming them. Thus reducing a well-meaning government initiative to gamesmanship.

Kudos to the EPA for calling VW out on this. This kind of incident reinforces (in me at least) the idea that environmental legislation is an extraordinarily difficult thing to get right.

[0] https://en.wikipedia.org/wiki/Goodhart%27s_law

[1] https://www.law.cornell.edu/wex/fiduciary_duty

[2] Catalytic convertors

>This kind of incident reinforces (in me at least) the idea that environmental legislation is an extraordinarily difficult thing to get right.

And the more detailed you make the regulation to prevent gaming, the harder it is for new competitors to enter the industry. Also good god having read and interpet their frequently self contradicting regs is a nightmare. Hell, Tesla got hit with an EPA penalty early on because they didn't certify the roadster's non-existant emissions.

To me the tests have to be real measurements rather than a proxy.

Having a special test makes the test easy to pass.

If the test was instead to have sampled data from x,000 cars, gathered at service intervals fit within certain bands, then that would require actually engineering most cars to return within the set range across a wide range of use types.

Smaller niche manufacturers could just pass a simpler test to encourage market entrants. 200 slightly worse cars is nothing against 200,000 intentionally bad cars.

The regulations need to target what is trying to be achieved rather than simplified proxies.

In this case, VW didn't game the regulation. They gamed the test conditions. They deliberately violated the regulation. Violations should be expected on occasion. Violators should expect to be punished when caught.

Seems very UnGerman to patch over a flaw instead of rooting out the problem from the ground level and engineering a solution. I'm just saying deceitfully patching over the problem seems uncharacteristic of the cultural norm/stereotype that I know.

How anybody at VW would think it's a good idea? Sure, it's a clever trick, but when they get caught - and it's assured they will - they'd probably get it pretty hard. I mean, Toyota got hit for $1.2 bln for problems that were mostly imaginary[1], how much this would cost?

[1] http://www.cato.org/publications/commentary/justice-departme...

If these controllers were open source, drivers could decide on their engine settings themselves. That could be an interestimg strategy for VW as it would move responsibility to the drivers.

It's my understanding that there's plenty of aftermarket opportunities available for this kind of tuning.

VW denies you warranty repairs if you do this tuning if the damage can be traced to ECU modifications. Class action lawsuit in 3...2...1...

That's a perfectly reasonable policy.

There is also quite some lies "optimism" in the EU regarding stated mpg L/100km from the manufacturer and real world measured mpg.

Is it the same software that the car companies use during EU drive cycle tests vs real world production cars? If you were a cheating manufacturer you would tune your car lean for the tests and more sporty with slightly higher consumption for the real world production run.

Sure, the real challenge must lie in quickly recognizing the different test setups for consumption vs pollution, because they certainly "require" very different engine parameters. I would not be surprised if (or rather: I'd be surprised if not) consulting industry experts managed to sneak some easily detectable patterns into the initial part of the standardized "test scripts". It's not like these kinds of optimizations are in any way brand exclusive, or it was impossible to know of them before these news (the only way to not know about them, assuming a moderate interest in all thing car related, was deliberately turning a blind eye for the sake of believing in progress)

It's hard to believe anyone still fooled by VW's claim of 94.2 MPG (1.6 TDI BlueMotion Golf, manual, extra urban). There's been a lot of press about stated fuel figures not matching real world consumption (and it's even mention/stated briefly in adverts now I think)

In my own experience with that model, driving about 500 miles, 50-55 is a valid expectation

That's okay. They just need to grow the wheelbase a bit and reclassify all their cars as "light trucks". Problem solved.

I'm wondering how this is going to be implemented.. it doesn't seem like a thing that's really fixable here.

Total speculation based solely on experience as a consumer (and I own a 2013 VW Golf TDI, so it affects me), but my guess is that the fix would be updated software to apply the stricter emissions controls all the time, presumably at the expense of engine performance.

EDIT: And reducing fuel economy, which is a huge selling point of the TDI in the first place. No idea "by how much" though, and typically TDI numbers are understated, so its possible that a reduction in economy will still be in line with their claims.

Yea, I have a 2015 Jetta which is why I was curious. I wonder if this explains why my car sounds really weird for the first few minutes of running.

It's probably the secondary air pump (sometimes nicknamed the "Good Morning Pump" by VW owners) that operates when the engine is cold, to reduce pollution https://en.wikipedia.org/wiki/Secondary_air_injection

I think that's just a diesel thing.

I don't have a diesel, but it sure sounds like I do!

This is interesting. VW has already been performing a nationwide ECU reflash on all MkIV TDI cars this year. Was that part of this action, or was it the cause of this action?

Given the low cost and ubiquity of sensors nowadays, you'd think it would be feasible to require car makers to test emissions in real time, in the real world.

We already have that, It's called OBDII.

Besides error codes, OBDII has 8 emissions monitors. In California all cars after 2001 are checked by means of a visual and these 8 readiness monitors: catalyst, heated catalyst, EVAP, SAI, O2, O2 Heater, EGR/VVT, AC

Source: I work on my two Audis

If you need to, you can beat most O2 monitors with a 555, most O2 heater monitors with a $0.45 resistor (or an O2 heater from a neutered O2 sensor), an EGR sensor monitor with a diode and couple resistors. Many catalyst monitors are just downstream O2 sensors that detect cycling (another 555 will defeat).

OBD2 monitoring, without fairly careful visual inspection, isn't a full solution. It's used because it's cheap, easy, and close to good enough, but enthusiasts are easily able to defeat it, for "off road use only", of course...

Until recently O2 sensors were not used on diesels. There are a host of reasons for this, one of the main ones being that until recently, diesel fuel in most places had high levels of sulfur, which is a known poison for Pt and some of the other metals which are used in both catalysts and O2 sensors.

As for EGR monitoring, most OBD2 systems I've seen monitor the presence of the egr solenoid by looking for the back emf. So you'd need an inductor or a solenoid for that. Additionally, they usually will often do a test of the manifold pressure while cycling the valve, to confirm the flow rate (obviously this can only be done of the engine in question has a MAP sensor). And if the MAP sensor is the primary sensor for doing fuel monitoring (ie, no MAF sensor), then there would be no easy way of defeating this monitor without messing up the normal performance of the engine. So in my experience, defeating EGR and passing the EGR related OBD2 monitors is non trivial. But then, since there is usually an allowance (in the current CA smog check schemes) for some of the monitors to not have completed running while still passing the test, it doesn't matter too much in practice (can't disconnect the EGR valve, but the egr readiness monitor is moot). The EGR monitor is usually the last one to run anyway, so this makes passing the smog check with what would otherwise be a egr related OBD2 failure, fairly easy.

Beating O2 monitors is actually much more difficult in practice. First of all, an increasing number of O2 sensors in cars are wideband sensors which don't output the traditional switching signal one could simulate with a 555. Even some older cars use wideband O2 sensors (I have personal experience with two MY 1998 Toyotas [california cars] for which this is the case). Secondly, even if you do have all narrowband O2 sensors on a car, the way the O2 monitors work is that the downstream (of the catalyst) O2 sensor(s) output is compared to the upstream (of the catalyst) O2 sensor(s) to see, not whether the O2 sensor is working, but that the catalyst efficiency is above some threshold. Simply simulating one or both signals with a naive switching circuit will trip the catalyst efficiency monitor. And why would you want to disable the upstream O2 sensor(s) anyway, it's main job is to provide an error signal to the fuel system which would otherwise be a completely open loop system? Without a working feedback loop your fuel system won't be running very well, and usually falls back to a super conservative fuel mapping because of the danger of unknowingly running lean (obviously a concern for gas engines only, not diesels).

Defeating OBD2 monitoring in these sorts of ways is not an issue in any meaningful sense. Enthusiasts often will replace their ECUs wholesale with an aftermarket one. When the ecu hardware and its programming is completely outside of the manufacturer's control, who can say anything about the validity of it's outputs for emissions compliance purposes?

I think he means to have the car collect it's own data on emissions, which can then be collected.

It's a far better idea than to just sample a few cars in a limited test.

2009-2015, I'm surprised it's only 1/2 million vehicles.

It's only the diesel models, which are still only a small fraction of the cars being sold today.

I'm keeping my old Mercedes diesels with no computers at all.

That's OK, I probably use more clean air than I need anyway.

Oh, you have a Unimog too?

The technical debt of petrol engines is becoming apparent.

The technical debt of petrol engines is not relevant to this article:

> The allegations cover roughly 482,000 diesel passenger cars sold in the United States since 2009.

Sorry, I'm not a native speaker. I thought the word "petrol" includes diesel, too!

No worries! "Internal combustion engines" might be the best phrase to use, then.

This reminds me of gaming hardware benchmarks.

I can't believe such a large company would cheat like that. I think there must be more to this. Maybe a rogue employee?

More like a Rogue CEO and Board. Decisions of this magnitude going over so many years likely were made by the people at the top.

Recall that VW Chairman Ferdinand Piech resigned earlier this year http://uk.reuters.com/article/2015/04/25/uk-volkswagen-ceo-c...

The board was quoted: "The members of the steering committee came to a consensus that, in the light of the past weeks, the mutual trust necessary for successful cooperation was no longer there"

What likely happened was that the EPA filed suit against VW in April, and in uncovering the evidence needed to support the suit, VW chair Ferdinand knew that he needed to go in order to let VW continue to be a company. Or, Ferdinand could have been the ethical one and blew the whistle.

Either way, let out the stink. Corporations aren't nearly the kind of flawless entities you think they are.

Corporations aren't flawless, but it definitely isn't a given that the people at the top have any clue what is going on with their software either. I can think of two instances that I've personally seen very similar to the referenced story where someone in middle to middle-upper management made the call, and only 1-2 devs/analysts had any idea what was done.

If anything that should go to show the lack of sophistication a lot of these projects have. There often is no project manager, no code reviews, and a messy codebase where snippets of very unethical logic can hide for years.

I can't believe such a large company would cheat like that.

Companies, at the time when "company" was still being defined, used to do anything to exploit people and kill the environment just so their owners got a return on investment.

Only through the past 100 years of laws, regulations, and cultural refinement have corporations overtly stopped so many abuses of power. Some of the refinement is purely psychological manipulation as well ("Clean Coal!") and companies are still doing awful things to individuals, societies, and the world at large.

Our current generation is blind to many of these abuses because we grew up inside the bubble of "friendly companies." Friendly companies know they have to maintain their image. Friendly companies know they must hide their evil deeds from law/regulation because public relations now matters more than beating people with sticks to increase their productive output (because companies are immortal after all, you can't have the public turning on you).

A large corporation...acting unethically? Say it ain't so! This is why regulations exist in the first place, because you can't trust companies to act ethnically on their own. Their job is to maximize profit, and the invisible hand only works on money.

> because you can't trust companies to act ethnically on their own

Indeed, this doesn't sound very German

You would think that they would be smart enough to only cheat when they won't get caught.

Here, they got caught, so they were being stupid.

You don't expect big, successful companies to be stupid (after all, they spend a ton of money to get the best CEOs they can).

GM knowingly declined to fix the safety problems with their ignition switches. This really killed people, not just violated some arbitrary local emissions number.

You are so right. GM didn't just kill a few people, they killed over 120 people:

   General Motors agreed to pay $900 million as
   part of a Justice Department investigation
   into its failure to fix a deadly ignition
   switch defect blamed for more than 120 deaths.

That is a shockingly high number. http://www.usatoday.com/story/money/cars/2015/09/17/gm-justi...

Poe's Law is in full effect, here...

While this seems to be made on purpose, my gut feeling is that car makers shouldn't do software...

Who should write the software then? People who don't know anything about cars?

People who know software, sitting in the same room as people who know cars.

I fear a big company would find it impossible, but that's how you get good software.

Are you under the impression they don't hire programmers to write their software? I don't understand your point.

Funny coincidence, reading the article we see that the american gov just punishes foreign companies.

What is coinciding with what?

DDOS VW shall we?

Does HN not detect duplicates? I submitted the same article 3 minutes before OP.

https://news.ycombinator.com/item?id=10239985

There's also another submission 15 minutes after OP's submission too.

https://news.ycombinator.com/item?id=10240106

It does to a point.

The OP's link after .html: ?_r=0

Your link after .html: ?module=Notification&version=BreakingNews&region=FixedTop&action=Click&contentCollection=BreakingNews&contentID=50913528&pgtype=Homepage

The 3rd link after .html: ?referrer=

In this case, while the links may lead to the same article, they are NOT the SAME link.

edited: to make url differences more visible

The URLs are slightly different, this post uses: http://www.nytimes.com/2015/09/19/business/volkswagen-is-ord...

Your submission uses: http://www.nytimes.com/2015/09/19/business/volkswagen-is-ord...

The urls are different for the one you submitted and OPs. Generally if you can remove extra query string params in the submission it helps the filter / deduper work better.

As I understand it, the new policy at HN is to allow duplicate posts to make sure enough readers see the article. I used to check for duplicates before posting, but now I don't bother.

Is this an example of software becoming a hindrance or more beneficial to the industry?

Good for Volkswagen.

I love the look/feel/drive of a VW, but I will never again own one.

My parents bought a 2002 VW which I mostly drove and the maintenance costs were through the roof. There have been at least 3-4 recall notices, one time while parked on an incline, it rained and flooded the back of the car with water. When I went to VW, they said that I had to manually clean out some drainage underneath the car. How the hell am I going to do that?

Then the coolant leaks, engine belts, cracked radiator. You name it I probably had to get it fixed.

Never again.

Interesting, since here in EU the VAG group cars (Seat, Škoda, VW, Audi) are widely perceived as the most reliable and there's a huge amount of people buying nothing else.

Perhaps it's where they are built. I assume most US-bound VWs are from Mexico or US proper due to NAFTA and auto-import regulations.

I had a 2009 Wolfsburg Jetta 2.0t, assembled in Mexico I believe, that was a complete maintenance nightmare. I got out of it before the timing belt needed to be replaced and purchased a Toyota instead.

My 2011 Jetta was built in Hamburg, according to the VIN plate.

Yes, VWs made outside of Germany are basically junk.

As was my German-built Corrado in the 90s. Of course, is that stopping me from buying another one this year? Nope, it just took me 15 years ago to get over it.

This was true for much of the 2000s, but the new platform used from ~2009 onwards (the one implicated here) has been perfectly fine. The 2009-2016 Golfs get consistent Very Good/Good reliability ratings from Consumer Reports.

They're still not Honda, but they're now better than average.

VW have a much better reputation than Honda in Europe.

The most common complaint is that repairs for German cars are much more expensive, but this may be an artifact of the supply chain for parts and not apply to Europe.

Individual anecdotes don't mean much. You can find objective data on car reliability on TrueDelta and Long-Term Quality Index. Those aren't true random samples but still pretty useful.

http://www.truedelta.com/car-reliability

http://tradeinqualityindex.com/reports/Volkswagen.html

_My_ story is anecdotal but early 2000 VWs are known for being really shitty. this have this history. I heard they have gotten "better" but a bit too late for me to care about.

I had a 2001 Passat 1.8T (that I got used in 2003) had a slew of problem covered under the VW CPO warranty. Fortuitously enough, after that warranty expired, it ran like a clock for 8 years. (But I wasn't afraid to service it myself either.

On the newer ones the high pressure fuel pump disintegrates and poops metal bits into the engine, requiring the engine to be replaced. That's why VW recently extended the warranty against the HPFP to, if I recall correctly, 100k miles.

Here's the 400-page-long forum thread on that defect: http://forums.tdiclub.com/showthread.php?t=308323

Note that the pump in question has been redesigned multiple times by Bosch. Hardware bugs are a thing, even in cars.

Maybe the bug is that the engine requires ludicrous fuel pressure to achieve its claimed performance.

IIRC, 2002 was around a local minimum in VW reliability.

Instead of fining the manufacturer, just fine everyone who owns one. If every VW owner was forced to pay 20k or so, the problem would fix itself.

Well, sure, if auto purchasers were at risk of massive fines if they received a vehicle where the manufacturer had done something that the average user wouldn't have the skill to detect, no one subject to the government imposing such fines would ever buy a car again, which would solve the problem quite completely.

OTOH, destroying the entire auto industry in the subject jurisdiction might not be considered an optimal means of resolving the problem. (And, indeed, might be so suboptimal to the population as to have catastrophic political consequences for the government imposing it.)

> destroying the entire auto industry in the subject jurisdiction might not be considered an optimal means of resolving the problem.

I don't see how passing the costs along to future car purchasers is better than extracting the costs from the people who were actually harming others.

I'm not sure if you are satirical.

How would fining the end users solve the problem?

Because no one will ever buy a VW again, and so other carmakers will take note.

And people will be on the hook for $20,000 because they purchased something from a now-nonexistent company?

That's idiotic.

Would anyone buy any vehicle ever again, given the possibility of a 20k liability for something out of their control?

This position has not been thought through.

Reply on news.ycombinator.com