EPA says Volkswagen installed software to cheat on emissions

[nsaspook]

Cut to about 22:00 for the VW discussion.

 

[256bits]

Bosch knew about it as early as 2007, when they wrote a letter to VW. According to Bild am Sonntag, Bosch wrote the software for testing purposes. In 2007 Bosch wrote a letter to VW, telling them that the use that VW intended was against the law. According to Bild, this letter was adressed to the highest management circles.
http://www.bild.de/geld/wirtschaft/...1-vor-abgas-manipulationen-42736218.bild.html (in German)

Thanks PeitKuip,
That's what I was wondering about. How much did Bosch know about the whole code being used in production.

 

[nsaspook]

Thanks PeitKuip,
That's what I was wondering about. How much did Bosch know about the whole code being used in production.

That's something that IMO will be hard to find out unless this ends up in a count with the power to break confidentially agreements under oath. Bosch looks like the hardware integrator for the ECM using the infineon TCL1796. If VW specified something like a special code protection module, drivers and demonstration code to hide firmware from prying eyes that would be nothing special as most advanced controllers have this capability in the hardware as options but if Bosch provided technical advice with the specific details and information of designing a defeat device with possible cloaking or cryptographic masking of the functions for the willing purpose of committing a fraud they would be in deep, deep trouble.

Most auto engine management software systems are networked with CAN bus so the information needed for the system to tell which engine mode it needed to operate in could have easily been linked in from several other modules each with a little part of the pie making it very hard to nail down exactly who did what if the architect of the system was on the ball.

Things like this are why the 'IOT' where networked computers are in everything might not be so great an idea in reality.

 

[edward]

My son bought a 2011 Passat TDI about three months ago. He had been waiting for quite a while to find a low mileage used one. He just checked with various sources and they tell him his vehicle lost over half of it's resale value. New car dealerships, not just VW, won't even take them as a trade in. The irony is that he loves the car just the way it is. It is really a quick little sedan that gets over 40 mpg on the highway.

The "treadmill" :) emissions test is a joke compared to real world driving. Very few cars would do as well out on the road. The test drivers creep up to speed very gradually.

If VW comes up with a performance and economy degrading fix how long will it be until the aftermarket comes out with something to defeat the fix?

 

[nsaspook]

If VW comes up with a performance and economy degrading fix how long will it be until the aftermarket comes out with something to defeat the fix?

I'm sure somebody will try but I would not like to the first one marketing the defeat 'fix' under this kind of heat.

 

[Astronuc]

The "treadmill" :) emissions test is a joke compared to real world driving. Very few cars would do as well out on the road. The test drivers creep up to speed very gradually.

My first Honda Civic (5 speed, standard) did better than the test mileage on the highway.

 

[nsaspook]

The Bosch system VW and many others use.

 

[nsaspook]

Then
http://www.volkswagenag.com/content/vwcorp/info_center/en/news/2015/09/sustain.html

Full marks were awarded to Volkswagen in the areas of codes of conduct, compliance and anti-corruption as well as innovation management, climate strategy and life cycle assessment. The Group is also the industry benchmark for supplier management and environmental reporting. Furthermore, significant progress has been made in human capital development, occupational health and safety, tax strategy and talent attraction.

Now
http://www.sustainability-indices.com/images/150929-statement-vw-exclusion_vdef.pdf

"Volkswagen AG to be Removed from the Dow Jones Sustainability Indices"

http://www.theguardian.com/environm...ars-been-unfairly-demonised-for-air-pollution

The car industry on Wednesday launched a campaign to “challenge the increasing demonisation of diesel” vehicles.

 

[edward]

It looks like Dow Jones is coming down really hard on VW. VW is a very big company and employs a lot of people. At this point I am a bit concerned about unintended consequences in a shaky global economy. S&P ?? how do they fit into this . S&P was one of the companies that gave packaged sub prime loans an AAA rating.

 

[nsaspook]

“The purpose seems to be to pass the peak luminance measurement test and then reduce luminance (and power) to get a better energy label ranking when the on power is measured,” the correspondence says. “All very clever and it is not dimming so much that it makes a huge difference, but does the commission consider this an acceptable practice or is this a non-compliant activity?”
...
More testing is planned to establish whether manufacturers are gaming television testing procedures. But “it wouldn’t take much for an unscrupulous manufacturer to install software to detect the unique ‘signature’ of the test and to then have the unit go into some sort of eco-mode and produce superior results (ie lower energy use) that wouldn’t occur under normal usage,” Horowitz said.

http://www.theguardian.com/environm...e-energy-efficient-in-tests-than-in-real-life

Now it's TV defeat devices.

https://www.schneier.com/blog/archives/2015/09/volkswagen_and_.html

Computer-security experts believe that intelligence agencies have been doing this sort of thing for years, both with the consent of the software developers and surreptitiously.

This problem won't be solved through computer security as we normally think of it. Conventional computer security is designed to prevent outside hackers from breaking into your computers and networks. The car analog would be security software that prevented an owner from tweaking his own engine to run faster but in the process emit more pollutants. What we need to contend with is a very different threat: malfeasance programmed in at the design stage.

 

[nsaspook]

It's interesting to look back with the current state of knowledge and see the comments and theories about the TDI cars in 2009 when people discovered the EPA ratings were low when compared to 'real world' driving.

http://www.greencarreports.com/news/1019256_volkswagen-jetta-tdi-much-more-mileage-than-epa-admits

His results confirm consistent reports from actual drivers that the EPA's official mileage numbers for the JettaTDI (29 mpg city / 40 mpg highway for the automatic, 30 / 41 for the manual) are far too low.

Last July, Volkswagen hired independent tester AMCI to test the Jetta TDI's"real world" mileage on the road. They came back with 38 city / 44 highway--or 24 percent and 10 percent higher respectively.

 

[Hornbein]

It's interesting to look back with the current state of knowledge and see the comments and theories about the TDI cars in 2009 when people discovered the EPA ratings were low when compared to 'real world' driving.

http://www.greencarreports.com/news/1019256_volkswagen-jetta-tdi-much-more-mileage-than-epa-admits

My cousin Scott informs me he was told about the VW trickery back in 2013.

 

[nsaspook]

My cousin Scott informs me he was told about the VW trickery back in 2013.

Are you sure it wasn't your cousin Vinny.

 

[Hornbein]

Are you sure it wasn't your cousin Vinny.

My cousin is Scott Howard, CEO and owner of Shandam Consulting. http://www.shandam.com/

 

[256bits]

Things like this are why the 'IOT' where networked computers are in everything might not be so great an idea in reality

IOT - too many systems to keep track of them all in the code.

I had this thought that the code design team got confused about which code was actually the production code, and inadvertently switched the test code to the running code, and visa-versa, with both sets being in there as a design choice.

The test code was to have the engine run raw ( so to speak ) and then add in systems or remove them to probe a problem. Satisfaction was when the engine performance and emissions became equal to design standards. Then the run code was set, and compared, and if equal again, the car was good to go, otherwise re-diagnose.

They could have unwittingly duped themselves, rather than deliberately duping the public.

That would be part of the quality control lax that was mentioned earlier ( by Borg I think.)

 

[Hornbein]

My cousin is Scott Howard, CEO and owner of Shandam Consulting. http://www.shandam.com/

Here's the relevant part of a letter he wrote to me.

I have a friend that works for the Bureau of Automotive Repair here in [Sacramento] CA – he said they picked up on the VW testing inconsistencies several years ago, and it took some time to bubble upward.

 

[nsaspook]

IOT - too many systems to keep track of them all in the code.

I had this thought that the code design team got confused about which code was actually the production code, and inadvertently switched the test code to the running code, and visa-versa, with both sets being in there as a design choice.

The test code was to have the engine run raw ( so to speak ) and then add in systems or remove them to probe a problem. Satisfaction was when the engine performance and emissions became equal to design standards. Then the run code was set, and compared, and if equal again, the car was good to go, otherwise re-diagnose.

They could have unwittingly duped themselves, rather than deliberately duping the public.

That would be part of the quality control lax that was mentioned earlier ( by Borg I think.)

You don't really believe that do you? They have flat out admitted to cheating. If there was the tiniest possibility of a coding mix-up that lasted for 6 years with many model revisions they would have the guys responsible for it frog-marched in from of the press in a nanosecond to save billions of dollars.

 

[mheslep]

There was no, is no, version of the software that makes the VW diesel "good to go". There's a version (mode) with good acceleration and mileage performance and emissions well over limit, and there's a mode with lousy performance but certifiable emissions. There's no mode that does both.

 

[nsaspook]

There was no, is no, version of the software that makes the VW diesel "good to go". There's a version (mode) with good acceleration and mileage performance and emissions well over limit, and there's a mode with lousy performance but certifiable emissions. There's no mode that does both.

As is usually the case a new version of this has arrived on the web: NSFW language
http://jalopnik.com/hitler-is-understandably-pissed-about-volkswagens-diese-1731943072

 

[Vanadium 50]

There's no mode that does both.

There is clearly no present-day VW mode that does both. I am less certain that there is no aftermarket chip that does both. There is a counter-culture of "chippers" who use aftermarket ECU software to improve performance. It would not strike me as impossible that one could get at or near design performance by using this modified software. Why doesn't VW do it then? More wear on various engine elements. But given a choice between replacing 500,000 cars now or 10,000 turbochargers over the next decade, which would VW prefer?

 

[russ_watters]

There is clearly no present-day VW mode that does both. I am less certain that there is no aftermarket chip that does both.

I am: the separate modes are mutually exclusive. You can have chemistry or physics, but not both at the same time. That's why the fraud was needed.

More wear on various engine elements. But given a choice between replacing 500,000 cars now or 10,000 turbochargers over the next decade, which would VW prefer?

Ok, no mode that does both without destroying itself or needing a refill of the hidden urea tank (guess) once a week. Same diff.

 

[Vanadium 50]

The ECU design balances four priorities: emissions, economy, reliability, and performance. VW decided they wanted economy, reliability, and performance at the cost of emissions. No reason why one couldn't tilt the balance differently: emissions, economy, and performance at the cost of reliability. Whether this is feasible is a matter of numbers: if you could get legal emissions and substantially similar economy and performance at a cost of reducing the average engine lifetime from 400,000 miles (I made this number up) to 399,999 miles, we would also say "do it!". If it changed from 400,000 miles to 1 mile, we'd all say that's ridiculous.

I'm not an expert on these chips, but expect that there is substantial additional wear on the EGR valve ($200-300 to replace) and some additional wear on the turbos themselves ($1500-$3500 to replace). If VW had to replace every EGR valve and 10% of the turbos over the lifetimes of these cars, they'd jump at the chance.

 

[mheslep]

It seems to me government can best serve consumers and the environment in this case by leaving the VWs already on the road alone and beating the heck out of VW corporate, thus sending a message to other mfns who might attempt a similar fraud. Air pollution is effected by the level of NOx emissions from the entire US fleet. The relevant VW diesels in the US start with 2009 models and appear to number in the thousands (482,000 ?). They really don't matter among a fleet of 200 million vehicles, IF the price for such fraud is shown to be extreme.

 

[russ_watters]

The ECU design balances four priorities: emissions, economy, reliability, and performance. VW decided they wanted economy, reliability, and performance at the cost of emissions. No reason why one couldn't tilt the balance differently: emissions, economy, and performance at the cost of reliability. Whether this is feasible is a matter of numbers...

Assuming it is even physically possible to tip the trade-off that way, sure: the car has a certain warranty (arrived at via competition) and needs a certain reliability for that to be economical. That's pretty much non-negotiable. It's the least malleable of the parameters.

I'm not an expert on these chips, but expect that there is substantial additional wear on the EGR valve ($200-300 to replace) and some additional wear on the turbos themselves ($1500-$3500 to replace). If VW had to replace every EGR valve and 10% of the turbos over the lifetimes of these cars, they'd jump at the chance.

I'm not either and I'm still unsure of how they are able to operate in a mode that passes emissions at all, given that other cars need a consumable catalyst, which these apparently do not have. That's why I'm saying I don't know that the trade-off you are suggesting is even possible. But sure, if it were a fairly minor reliability issue, I'm sure they'd jump at the chance. The fact that they chose fraud instead implies to me that it is worse than that.

 

[Vanadium 50]

Well, we know that they balked at a $300 urea-based NOx remover, so a reliability reduction of $300 equivalence would also be rejected. Rejecting that putative tune point may have made sense in 2007. It may look a lot more reasonable today, when faced with fines that are two orders of magnitude larger.

Looking at dyno curves, it looks like there is about a 10% loss in power in test mode. This is a number that is not out of the question to get by reprogramming ECUs. Maybe the EGR valve dies sooner.

 

[mheslep]

A 10% power loss by choice is out of the question for most owners I suspect.

 

[Vanadium 50]

The loss is 10% if you do nothing beyond . I believe that may well be recoverable by programming. 20% or so can be gained by increasing the turbo boost, which will, in turn put more stress on various systems. That may reduce reliability, but as I argued above, this may be the least expensive alternative for VW.

 

[nsaspook]

Any 'fix' must include another one to fix the corrupt executive management and corporate structure.

 

[nsaspook]

http://www.latimes.com/business/autos/la-fi-hy-vw-cancels-2016-diesel-line-up-20151007-story.html
http://docs.house.gov/meetings/IF/IF02/20151008/104046/HHRG-114-IF02-Wstate-HornM-20151008.pdf

CSPAN schedule: http://www.c-span.org/video/?328599-1/hearing-volkswagen-emissions-violations

The decision came to light late Wednesday afternoon after Volkswagen's U.S. chief executive, Michael Horn, released written copies of testimony he is expected to give before Congress on Thursday.

In the testimony, in which Horn offers "a sincere apology for Volkswagen's use of a program that served to defeat" emissions tests, the executive said, "We have withdrawn the application for certification of our model year 2016 vehicles.
...
"They’ve abandoned the entire 2016 model year (2.0L diesel engine) diesels, and that's not good news for owners," Brauer said. "It suggests that the fix is probably not going to be easy. It suggests that the fix involves so much challenge that they’re not even going try to make the 2016s work."
...
Update: The EPA issued a statement saying, "Today Volkswagen withdrew their certification application for 2016 vehicle models that use the 2.0L diesel engine including the AUDI: A3 VOLKSWAGEN: BEETLE, BEETLE CONVERTIBLE, GOLF, GOLF SPORTWAGEN, JETTA, PASSAT models."
A Volkswagen representative said the Touareg diesel, which uses a 3.0-liter engine, was not affected by the EPA discussions and would be part of the 2016 line-up.

 

[nsaspook]

The hearing is on CSPAN3 online now.

Quick notes from the hearing:

No software only fix for at least the 1st gen cars 430,000 cars.
Needs AdBlue and catalytic converter, no current time frame for a fix.
Fix will affect overall drive-ability, possible compensation to customers of affected cars.

Software and some hardware senors added ,maybe next year for 2nd gen cars, 95,000 cars.

Dealers: VW to fund money to save dealers. No plans to buy back current new car inventory.

 

[nsaspook]

Find this in the hearing.
The statements from rep. Chris Collins (R-NY and a engineer) on VW cheating were right on point.

 

[nsaspook]

http://www.kcci.com/project-economy/volkswagen-faces-tough-questions-on-capitol-hill/35722356

"This was not a corporate decision. To my best knowledge today, this was a couple of software engineers who put this in for whatever reason," said Michael Horn in response to questions from Rep. Joe Barton during a Congressional committee on the scandal.

Barton expressed disbelief with Horn's answer, saying he couldn't believe that no one in management was aware of such an important decision considering "how well run as VW has always been."

"I agree, it's very hard to believe," Horn said.
...
Rep. Jan Schakowsky, an Illinois Democrat, said that Volkswagen should buy back the diesel cars at the original purchase price. "If they want it, every VW clean diesel owner should be able to get their money back," she said. Horn said the company is looking at how to compensate owners but that its plans are to fix the cars, not repurchase them.

Schakowsky said assurances from the company that the cars will eventually be fixed, perhaps by the end of next year, are not enough.

"Volkswagen's word isn't worth a dime," she said. "To find a company that deliberately cheated asking customers for patience, is not acceptable."

Following Schakowsky's criticism of VW, Rep. Chris Collins, a New York Republican, echoed her remarks.

"VW is trying to get the United States of America to believe this is the work of a couple of rogue engineers," he said, "and I categorically reject that."

I agree, it's very hard to believe.

 

[nsaspook]

The EPA say's it's doesn't understand or needs to understand exactly how VW cheated, only that they pass the test with the new fix. I find that very hard to believe if you want to expose what they did. They came across as easy to fool clowns in this hearing. No wonder VW thought they could and did fool them for so long.
Why, What, Where is damn important to a investigation. I would demand VW show me every line of code for the cheat device along with the people who wrote it before any negotiations about a 'fix' or they would never be able to sell a spoon in this country.

 

[nsaspook]

Video of the first part of today's hearing. The remarks from Rep. Chris Collins is about 1:48 into the video.
http://www.c-span.org/video/?328599-1/hearing-volkswagen-emissions-violations&live

 

[Astronuc]

Volkswagen Investigation Focus to Include Managers Who Turned a Blind Eye
http://www.nytimes.com/2015/10/26/b...-include-managers-who-turned-a-blind-eye.html

The number of Volkswagen executives or engineers suspended in connection with the emissions cheating has continued to grow and could soon reach about 10, said the person, who spoke on the condition of anonymity because the inquiry is supposed to be confidential. Some of the employees were directly involved in programming cars to cheat on emissions tests, but others may share blame because they found out about it and did not pass the information up the chain of command.

Video of the first part of today's hearing. The remarks from Rep. Chris Collins is about 1:48 into the video.

Chris Collins is an engineer!