Cars aren't only required to pass emissions tests, they're required to have their emissions within the specified limits at all times. The purpose of the tests is to ensure compliance.
Not true. Emissions vary a lot depending on environmental factors and optimizing for the worst case scenario (which is what "within limits at all times" implies) would likely result terrible averages. Even within the test cycle, the engine can be as dirty as they like at any point in time, as long as the average stays within bounds.
What they are not allowed to do is change the way the engine operates when under test vs on the road, and this is exactly what happened at Volkswagen. The VW case is not the first (ruling like this tends to be reactive, and slow...), but it might be the first were the response is firm enough to deter future would be offenders.
If it were the case, the regulations would say "No more than X NOX emitted per mile, even when climbing Mount Everest, even when 5 years old, even when poorly maintained, and even when used with low quality fuel, and in all other conditions anyone might put them in."
That clearly isn't reasonable, so instead the rules are written to say "They must pass XYZ tests". As long as those tests are passed, they meet the standard.
IMO, the main culprit here is the lawmakers for not correctly designing the rules and test procedures to be sufficiently realistic. Specifically, they shouldn't allow the CO2 test and NOX tests to be done in different conditions.
They should instead say "your car will be assigned to one member of EPA staff every day for a month. They will drive their regular daily journeys with it, and all emissions recorded. If you aren't happy with the test results (for example due to a cold week), you may ask for a retest.".
The problem with your reasoning is that these cars were designed to not meet the emissions requirements from the factory. From day one, they were only designed to meet the emissions standards during the testing, not on the road. They were purposefully designed to cheat on the test.
Your statement would be true if test performance varied from real-world performance due to environmental factors, but that's not the case. Test performance varies from real-world performance by design. And then they lied about it. That's why they're being punished.
I do agree that the test should be changed to take that inevitability out of the equation, but for this situation the reason they're being punished is because they designed their cars specifically to cheat on the existing test.
> Your statement would be true if test performance varied from real-world performance due to environmental factors, but that's not the case.
..or due to engine temperature, speed, gear, gas pedal position and tons of other variables WHICH IS the case. That's why emission requirements are (and must be) test-specific.
I'm sorry but you're wrong, and since I know you read my comment (you quoted it) it almost seems like you're deliberately wrong.
Yes, testing varies from the real world based on a number of factors. Environmental, driving style, etc. And the current tests don't care about that, which is a completely separate problem. The problem that VW got in trouble for is specifically designing their cars to cheat on the existing test.
Yes, cars behave differently if they're driven uphill vs downhill, if they're driven in the rain vs in the snow, if they're driven in the mountains or at sea level, but that doesn't matter at all. The test doesn't care, the law doesn't care. They're willing to accept that difference, as long as the car doesn't actively circumvent the purpose of the test. Which VW did.
You can argue that the test is invalid, and I agree. But that doesn't matter. The test is the test and VW cheated it. They built their cars specifically to have a completely different performance in the test vs in the real world with the specific intent to cheat the test.
Nope. When doing business, the only thing that counts is formal, verifiable spec. Crying "You have cheated since I wanted somehing else" should be irrelevant in any country that honors rule of law.
> specifically designing their cars to cheat on the existing test.
Test is a predicate. It can pass or fail, but you can't cheat it.
> actively circumvent the purpose of the test
Hmm, that's probably closest to something I can accept as an argument. Nice. But to circumvent the purpose requires an actual purpose. If the purpose is formal, it's the test itself. And if it's not formal, then it's IMO impossible or very hard to prove that it has been circumvented.
Laws are not software programs. Intent is a part of them, and it was clear the intent was for the testing to be representative of actual usage. It was not because the cars detected a test and changed behavior. That's a clear attempt to circumvent the regulation.
It's some time since I read the details but I remember that lots of people agreed it was not the actual usage, which btw made detecting it so easy. The acceleration was unrealistically long and slow, for example.
There is such thing as "evading the law" you know. It's not enough that you satisfy the regulation, if you act against the intent of the law in other ways. Or perhaps it's different in the US, but that's what I would expect in most of Europe.
It doesn't really matter what they're doing or saying now, what matters is what they were doing and saying at the time the evaded/avoided the law. If I admit to breaking the law, I still broke the law and I will still be punished for it. It doesn't retroactively make it okay just because I've admitted to it.
At the time they submitted their cars for testing, they told the world these cars met the test standards and did not say they had software to detect the test and modify the engine tuning. They only admitted it after they got caught, so at the time of the testing, they were evading the law, not avoiding it.
Avoiding the law would be not submitting their cars for testing.
