3 ms·

I'm not sure what your definition of alter is, but for me to alter means to lie. They weren't simply optimizing for test conditions, they were outright lying d

by rnovak 11y ago

I'm not sure what your definition of alter is, but for me to alter means to lie. They weren't simply optimizing for test conditions, they were outright lying during test conditions. that's what the EPA is saying is against the law. Here's the thing, and this is my main point: If the EPAs test conditions reflected normal driving habits, there would be zero difference between Manufacturer reported numbers and real world reported numbers. If the EPAs test is completely different from normal driving, then it stands to reason that there is no inherent correlation between manufacturer numbers and reported numbers. IDK if you've ever been through an emissions test (like the ones GM actually goes through), but they're a complete joke as to simulating real world behavior. Edit:Format

hinkley 11y ago
It's the old saw about the drunk looking under the light post for his keys, when he probably lost them in the alley; the light's better over here. If you make laws that can't be verified then you can't enforce them. So you test for things you think you can actually test and hope everything works out. Wiring up an exhaust sensor to everybody's car every time they need to renew their tabs and having them drive around the city is so infeasible as to be laughable. So they put your car on a dyno or even have you put it in neutral with the parking brake on and have you rev the engine a little bit and see if the tailpipe is a mess. Nevermind that a lot of cars belch smoke only under load.
semi-extrinsic 11y ago
So I threw together a histogram [1] of car CO2 emissions as reported by manufacturers for 1200+ model/engine combos. It's not exactly uniformly distributed, to say the least, there is a strong peak at 99 g/km etc. Now that's maybe not conclusive evidence that they are lying, but it's certainly suspicious. [1] http://asmunder.github.io/2015/09/A-histogram-of-CO2-emissions-as-reported-by-car-companies/ http://asmunder.github.io/2015/09/A-histogram-of-CO2-emissio...
- rnovak 11y ago
  I'm not sure what that proves, at least relative to what I'm saying? IFF the EPA's regulatory test procedure accurately described/expressed normal driving habits, then the only way that numbers would differ between the real world and the reported numbers would be fraud/lying/alteration, like what Volkswagen Group engaged in (let there be no mistake, it WAS fraud/lying, not a case of poor optimization). BUT the EPA's regulatory test DOES NOT accurately depict normal driving conditions. In the current case, a manufacturer has to: - Realize and Formally accept that the test procedure is inaccurate - Optimize for: - Simulated driving conditions, aka the Test Procedure - Real life driving To me, it looks like GM just didn't do any of that, and only optimized for the simulated conditions and not real world road conditions. Depending how far out of wack the test procedure really is could explain the entire data-set.
  - semi-extrinsic 11y ago
    So the point is that these official numbers should be taken with a generous pinch of salt. Now, when you talk about GM, remember this is data from Germany, so GM essentially means Opel. Now I believe that before the quite successfull Insignia which launched in 2008, Opel were mainly known for cheap and efficient and small cars like the Corsa and Astra, while now they have a more premium lineup, and "Turbo" versions of several models.
    - rnovak 11y ago
      Yes exactly, with a pinch of salt. Exactly what I was trying to say. And yeah, I've driven a Chevy Spark, and that thing was so slow I could get out and walk faster. I could totally see people driving with their foot to the floor in it.
joe_the_user 11y ago
IDK if you've ever been through an emissions test (like the ones GM actually goes through), but they're a complete joke as to simulating real world behavior. Does not matter. Your argument is something like, that by dumb luck auto manufacturers produced good emissions results and terrible results in the field. As understand it, the EPA is saying that if a manufacture is engaged in your two stage approach, "Optimize for:- Simulated driving conditions, aka the Test Procedure [and achieve high fuel economy] - Real life driving [and achieve high power, low fuel economy]" then they are engaging in systematic fraud and are going to get hit big time. You seem to think it's just a matter of technicalities in the test condition. I'd like references for that.
- rnovak 11y ago
  No. That's not my argument. It has absolutely nothing to do with dumb luck. The EPA has a set of standards, and manufacturers optimize for those. That's it. Cheating != optimization. Cheating is like a kid copying someone else's answers. Optimization is studying before the test. Completely different things. And optimization isn't illegal. If you still don't get it, I don't know what to say, I can't explain it any further.
  - joe_the_user 11y ago
    OK, I get that you saying "Cheating != optimization". That may be your opinion. It may be true in a moral sense. However, I have good evident that VW's optimization was considered criminal in this instance. And Indeed, since many manufacturers do this, things are pretty dicey. "“Developing an engine software to optimize certain aspects of an operation cycle that you know the parameters of is a challenge, but it is very possible,” says Thiruvengadam. “Knowing when to switch to the EPA-favorable cycle is the trick; it could be set up to detect the absence of steering-wheel movement, or, and this is known, we often turn off the traction control for testing purposes.” Either way, the result is the same: it turns the emissions controls on for EPA testing and off for real-world driving. Somewhat ironically, the presumed benefits of turning off the controls for normal driving include improved fuel economy and engine power." [1] [1] http://blog.caranddriver.com/how-volkswagen-got-busted-for-gaming-epa-diesel-emissions-standards/ http://blog.caranddriver.com/how-volkswagen-got-busted-for-g...

by rnovak 11y ago

hinkley 11y ago

It's the old saw about the drunk looking under the light post for his keys, when he probably lost them in the alley; the light's better over here. If you make laws that can't be verified then you can't enforce them. So you test for things you think you can actually test and hope everything works out. Wiring up an exhaust sensor to everybody's car every time they need to renew their tabs and having them drive around the city is so infeasible as to be laughable. So they put your car on a dyno or even have you put it in neutral with the parking brake on and have you rev the engine a little bit and see if the tailpipe is a mess. Nevermind that a lot of cars belch smoke only under load.

semi-extrinsic 11y ago

So I threw together a histogram [1] of car CO2 emissions as reported by manufacturers for 1200+ model/engine combos. It's not exactly uniformly distributed, to say the least, there is a strong peak at 99 g/km etc. Now that's maybe not conclusive evidence that they are lying, but it's certainly suspicious. [1] http://asmunder.github.io/2015/09/A-histogram-of-CO2-emissions-as-reported-by-car-companies/ http://asmunder.github.io/2015/09/A-histogram-of-CO2-emissio...

rnovak 11y ago
I'm not sure what that proves, at least relative to what I'm saying? IFF the EPA's regulatory test procedure accurately described/expressed normal driving habits, then the only way that numbers would differ between the real world and the reported numbers would be fraud/lying/alteration, like what Volkswagen Group engaged in (let there be no mistake, it WAS fraud/lying, not a case of poor optimization). BUT the EPA's regulatory test DOES NOT accurately depict normal driving conditions. In the current case, a manufacturer has to: - Realize and Formally accept that the test procedure is inaccurate - Optimize for: - Simulated driving conditions, aka the Test Procedure - Real life driving To me, it looks like GM just didn't do any of that, and only optimized for the simulated conditions and not real world road conditions. Depending how far out of wack the test procedure really is could explain the entire data-set.
- semi-extrinsic 11y ago
  So the point is that these official numbers should be taken with a generous pinch of salt. Now, when you talk about GM, remember this is data from Germany, so GM essentially means Opel. Now I believe that before the quite successfull Insignia which launched in 2008, Opel were mainly known for cheap and efficient and small cars like the Corsa and Astra, while now they have a more premium lineup, and "Turbo" versions of several models.
  - rnovak 11y ago
    Yes exactly, with a pinch of salt. Exactly what I was trying to say. And yeah, I've driven a Chevy Spark, and that thing was so slow I could get out and walk faster. I could totally see people driving with their foot to the floor in it.

joe_the_user 11y ago

IDK if you've ever been through an emissions test (like the ones GM actually goes through), but they're a complete joke as to simulating real world behavior. Does not matter. Your argument is something like, that by dumb luck auto manufacturers produced good emissions results and terrible results in the field. As understand it, the EPA is saying that if a manufacture is engaged in your two stage approach, "Optimize for:- Simulated driving conditions, aka the Test Procedure [and achieve high fuel economy] - Real life driving [and achieve high power, low fuel economy]" then they are engaging in systematic fraud and are going to get hit big time. You seem to think it's just a matter of technicalities in the test condition. I'd like references for that.

rnovak 11y ago
No. That's not my argument. It has absolutely nothing to do with dumb luck. The EPA has a set of standards, and manufacturers optimize for those. That's it. Cheating != optimization. Cheating is like a kid copying someone else's answers. Optimization is studying before the test. Completely different things. And optimization isn't illegal. If you still don't get it, I don't know what to say, I can't explain it any further.
- joe_the_user 11y ago
  OK, I get that you saying "Cheating != optimization". That may be your opinion. It may be true in a moral sense. However, I have good evident that VW's optimization was considered criminal in this instance. And Indeed, since many manufacturers do this, things are pretty dicey. "“Developing an engine software to optimize certain aspects of an operation cycle that you know the parameters of is a challenge, but it is very possible,” says Thiruvengadam. “Knowing when to switch to the EPA-favorable cycle is the trick; it could be set up to detect the absence of steering-wheel movement, or, and this is known, we often turn off the traction control for testing purposes.” Either way, the result is the same: it turns the emissions controls on for EPA testing and off for real-world driving. Somewhat ironically, the presumed benefits of turning off the controls for normal driving include improved fuel economy and engine power." [1] [1] http://blog.caranddriver.com/how-volkswagen-got-busted-for-gaming-epa-diesel-emissions-standards/ http://blog.caranddriver.com/how-volkswagen-got-busted-for-g...