What? Seriously? Volkswagen?

They seriously did this?

The origin of the case was a report last year by the International Council for Clean Transportation and West Virginia University that documented elevated emissions from some Volkswagen cars.
The cars in question could emit as much as 40 times the legal standard of nitrogen oxide, the EPA said.
When regulators initially raised the issue with Volkswagen, the automaker blamed the elevated pollution on “various technical issues and unexpected in-use conditions,” EPA said in the violation order.
Volkswagen in December last year initiated a voluntary recall of about 500,000 cars. Regulators broadened their probe when the cars continued to pump out excess emissions after the recall, despite showing some improvement, the California board said.
At that point, regulators told Volkswagen that they would not approve the automaker’s 2016 models “until VW could adequately explain the anomalous emissions and ensure the agencies that the 2016 model year vehicles would not have similar issues”, the EPA said.
“Only then did VW admit it had designed and installed a defeat device in these vehicles in the form of a sophisticated software algorithm that detected when a vehicle was undergoing emissions testing,” the EPA said.

Everyone knows that the mpg numbers and so on are, well, not a lie, but tweaked. Very careful drivers, driving in a very careful manner, are used to create them.

But actually having a specific engine mode purely for emissions testing? Which is only used during emissions testing?

That’s, umm, outrageous. And it’s not something I would normally believe. I’d, if presented with that story, say that the EPA has managed to get a bit mixed up there with the technology somehow. But they do seem very sure indeed.

Time to edge in on some of Julia’s popcorn futures perhaps? Because this could get very interesting indeed. Either way: whether VW really did do this or the EPA hasn’t quite grasped the point properly.

And there’s a further implication. It’s all about diesels, so few American manufacturers are going to be at risk here (they generally just don’t put diesels into consumer vehicles). But were the VW pollution stats anomalous as compared to other European vehicles? And if not, then who else has something similar installed?

From the EPA’s announcement:

In September, after EPA and CARB demanded an explanation for the identified emission problems, Volkswagen admitted that the cars contained defeat devices.

So, are they bang to rights?

15 thoughts on “What? Seriously? Volkswagen?”

  1. Everyone knows that the mpg numbers and so on are, well, not a lie, but tweaked. Very careful drivers, driving in a very careful manner, are used to create them.

    Not really – they are taken in a controlled, regulated environment. So, if they don’t reflect real-world operation, it is because the regulations & testing procedures are invalid. It is hard to see why car makers, who don’t (I assume) write the regulations, are to blame.

    But actually having a specific engine mode purely for emissions testing? Which is only used during emissions testing?

    That’s, umm, outrageous. And it’s not something I would normally believe.

    Nor me. I don’t know the testing procedure, but it is hard to see how you could get a sufficiently exact signal that testing was in progress to an engine management system reliably. Happy to be proven wrong though.

    It just sounds like, once again, testing procedures don’t test what they are supposed to test, and bear no relation to what happens in real life.

    Of course, any car maker would tune their vehicles to pass a test, however stupid it is, no doubt to the detriment of actual usage. When catalytic converters were made compulsory, lots of lean-burn technology was sidelined as it couldn’t work with them. So potential MPG savings were thrown away to satisfy emission regulations.

    It looks more likely to be incompetent regulation than anything else.

    Given the EPA’s recent track record, I would be inclined not to trust any of their pronouncements on principle anyway, until independently verified.

  2. I believe it. There’s been similar after market devices for a while, to switch between performance and pollution test modes. That they’d actually put it in at the factory… Someone’s getting fired. Or transferred to the crash test dummy department.

  3. I bought a new Bosch washing machine earlier this year. Among the various programs that may be selected are two that the instruction manual tells you very clearly are purely for energy efficiency testing and should not be used by consumers.

  4. Yet another argument in favour of open source software, I would have thought. Not everybody understands computer code but if everybody can see it there will be enough people who do understand it to make episodes like this even more unlikely.

  5. well, I used to put white kerosene in my Merc diesel before taking it off to Conneticut’s bi-annual pollution test … passed every time.

  6. “Everyone knows that the mpg numbers and so on are, well, not a lie, but tweaked. Very careful drivers, driving in a very careful manner, are used to create them.”

    Funnily enough, I was in the test cell block of a major lorry engine manufacturers this week, and was talking to them about how they develop their test cycles – basically they have a lorry with the same engine as they are testing rigged with a load of measuring gear, and a data logger. They take this out on a selection of test runs on the road, then programe the test cell to replicate the data from the road runs.
    Then they have a fully repeatable real world duty cycle they can play with without leaving their nice warm building.

    I suspect lorry MPG data is fairly accurate – truck opperators will pick up fairly quickly if they find their waggons are using much more fuel than advertised.

    Cars are another story – many manufactures program their in car MPG counters to read a bit better than reality, safe in the knowledge few owners will ever check…

  7. I’m only surprised there isn’t more of this about…or at least more noticeable.

    I can’t understand greenies with their energy saving shtick, it’s like they believe that we all want to burn fossil fuels and / or electricity for shits & giggles.

    Of course we want to use less, it would cost us less to run them! But we also want the bloody things to work properly!

  8. Yet another argument in favour of open source software, I would have thought

    It is always possible to ‘disassemble’ the compiled code, though its a bigger job than just looking at the source code.

  9. It’s just teaching to the test. If you know how the test is done, it’s easy to tell the ECU to switch modes. For example if the car is on a rolling road, you can tell by the fact that there’s no air rushing into the front of the car.

    I thought it was common knowledge that car manufacturers teach to the test? In fact there’s some ongoing debate about the Euro 6 emissions regulations, because some parts of the EU actually want to make the best better resemble real world conditions, whereas Germany quite likes the status quo.

  10. Car rolling roads usually have fan packs to give an air flow that matches (at least to a degree) the rolling speed, otherwise the cooler group doesn’t get enough airflow and the engine of whatever is on test boils up on short order if you thrash it.

  11. Years ago I discussed emissions control with an automotive engineer. They said that the controls are off for the first part of a driving cycle in order to give the catalytic converter time to warm up. There are special modes for testing on a rolling road in order to give representative figures for the vehicle’s emissions.

    So my guess: someone at the EPA doesn’t understand the engineering. Most of this is emissions control field is dominated by CARB anyway: the EPA should call those guys in. They do understand how cars work.

  12. If anyone thinks that testing to a product to standard means that every single item manufactured meets that standard then they have been fooled by the regulators. Testing is done on a representative item, not every single one, and many tests do not match real world use. I’ve been involved in IP & EMC testing and adjusting the product so that it meets the standard is common amongst all manufacturers.

Leave a Reply

Your email address will not be published. Required fields are marked *