Ford say their diesel engines meet EPA standards

[jpd80]

When you set up accredited independent labs, they can do work for multiple clients at differnt times

and it may be more profitable for auto companies to head that way.and pay fees per car for final validation.

 

You will never stop an out and out case of fraud but the government should move to close a few open gates,

I find it amazng that so many companies are permitted to self manage obligations to statutory certification.

 

Less opportunity to keep reults "in-hous" would restrict the opportunity to steer testing away from problem areas.

We saw this recently with Ford using an incorrect coast down value which made economy figures look better than they were.

 

..

Edited September 23, 2015 by jpd80

[RichardJensen]

I get all that, but this fraud would not have been prevented by independent testing or testing by the EPA.

 

The Hyundai/KIA fraud wouldn't have been prevented either.

Edited September 23, 2015 by RichardJensen

[jpd80]

I get all that, but this fraud would not have been prevented by independent testing or testing by the EPA.

That's because no testing outside the test loops is required,

I see the Europeans were moving in that direction (2017) but not quick enough.

An independent lab would in all likliness do additional tests because errors would be on them, not the manufacturer.

 

 

The Hyundai/KIA fraud wouldn't have been prevented either.

Hyundai's 40 mpg claim was never checked independently...

That 40 mpg claim was derived from the test cycles where Hyundai applied an incorrect coast down factor (like Ford)

 

Funny with manufacturer based tests, the errors always seem optimistic and favor the manufacturer

We'd struggle to find cases where the results were adjusted up because they were pessimistic...

 

Richard, I get your POV that it's impossible to close off all avenues to get around regulations and tests,

but that should not stop the government from trying to limit the potential for glaring errors getting through

and while it's good to see small independent reviewers exposing some "howlers", you'd think those easy

performed checks would be part of independent feedback on initial testing before certification...

Edited September 24, 2015 by jpd80

[RichardJensen]

That's because no testing outside the test loops is required,

I see the Europeans were moving in that direction (2017) but not quick enough.

An independent lab would in all likliness do additional tests because errors would be on them, not the manufacturer.

 

 

Hyundai's 40 mpg claim was never checked independently...

That 40 mpg claim was derived from the test cycles where Hyundai applied an incorrect coast down factor (like Ford)

 

Funny with manufacturer based tests, the errors always seem optimistic and favor the manufacturer

We'd struggle to find cases where the results were adjusted up because they were pessimistic...

 

Richard, I get your POV that it's impossible to close off all avenues to get around regulations and tests,

but that should not stop the government from trying to limit the potential for glaring errors getting through

and while it's good to see small independent reviewers exposing some "howlers", you'd think those easy

performed checks would be part of independent feedback on initial testing before certification...

 

The problem with testing outside the programmed loops is that if you drive outside the test parameters, you're not going to get comparable results.

 

Say I've got it in for VW, so I decide that I'm going to just mash the throttle and drive really aggressively on a "verification" test---well, surprise, surprise, it's not going to pass, because I'm driving outside the test parameters.

 

And if you have dozens of drivers performing a predictable 'independent' verification cycle, in order to get statistically valid results, then you've created another recognizable loophole.

 

---

 

The 'coastdown' factor that Ford & Hyundai supplied, used to calculate the rolling resistance on the dyno could've been gamed thusly:

 

EPA has to test pre-production vehicles, right? They can't buy a vehicle from a dealership in order to test it, right? So they have to trust that the pre-production model they get from the manufacturer is representative and not a ringer. So, either Hyundai does its own testing using a ringer, or they furnish a ringer to the EPA. Either way, you can easily game the system.

[RichardJensen]

Basically, if something has to be repeatable it can be rigged.

[SoonerLS]

Basically, if something has to be repeatable it can be rigged.

God only knows how many organizations have tested VWs over the last several years, yet only one of them noticed something not jiving.

 

At least we can be thankful for one thing--it wasn't those self-serving idiots at CR who found it...

Edited September 24, 2015 by SoonerLS

[ANTAUS]

[akirby]

You'd have to do emissions testing while driving which normally isn't at all necessary.But

 

It wouldn't be hard for the EPA to spot check though. Warm up the vehicle, take it on a normal drive for 30-45 minutes simulating the EPA test cycles as much as possible while measuring the tailpipe emissions and compare that to the lab results. It doesn't have to be exact, but if the real world result is an order of magnitude or more off then it bears further scrutiny. You could also establish a real world baseline from other vehicles - so let's say the real world testing on other vehicles is typically 10% or 20% higher then you look for anything outside of that.

[jpd80]

 

The problem with testing outside the programmed loops is that if you drive outside the test parameters, you're not going to get comparable results.

 

Say I've got it in for VW, so I decide that I'm going to just mash the throttle and drive really aggressively on a "verification" test---well, surprise, surprise, it's not going to pass, because I'm driving outside the test parameters.

Go back a step

I was suggesting that the labs do the test loops but also test outside those loops for anomalies like we've just witnessed.

Hook the test rig on the back of the cars and go for a test drive under normal conditions, the NOX is a result of efficient combustion

and will be worst at light crulsing, so the combined results would give confirmation that no defeat mechanisms are in play.

 

There's less chance of gaming the system, especially when those independent labs are also subject to auditing.

 

If you don'y try to change the system because someone can still try to game the system, then nothing will change,

start somewhere and evolve auditing systems until all bases are covered.

Edited September 24, 2015 by jpd80

[lfeg]

How about this - Since University of West Virginia did the initial testing (under contract) that brought this to light, the equipment needed to do on road spot checks exists. They and other entities could be retained to test randomly selected in service vehicles. With a good statistical selection method for test subjects, we could get a pretty good representation of the emissions state of the population of vehicles on the road without developing new procedures and facilities over a period of time.

[jpvbs]

We always see test mules packed with electronics. I'd guess they have emissions monitoring sensors in there for real time testing.

 

Which also makes me wonder if other manufacturers didn't test a VW to see what they were doing. We here all these stories about Mazda, Subaru, and others that were going to release a small diesel but keep delaying. You would think they would have reverse engineered a VW to figure out how they did it and in the process figured out they were cheating.

[RichardJensen]

Ok. How much money are you all prepared to spend to catch one particular kind of cheat?

[jasonj80]

God only knows how many organizations have tested VWs over the last several years, yet only one of them noticed something not jiving.

 

Personally I think the guy that found this was tipped off as to what to start to looking for by someone at an OEM. This all started almost 2 years ago with him investigating. An OEM doing calibration road testing would have noticed this in 5miles of real time driving with the NOx levels. I mean wasn't Mazda going to start selling a diesel here, made that announcement almost 2 1/2 years ago, Then delayed and delayed because they couldn't get the performance and emissions under control.... You had to think they would be benchmarking a vehicle already on the market that is the market leader they would be going after. OEM's never go after other OEM's for something like this but wouldn't have any steering someone in the direction.

[Biker16]

There's a bright line between criticizing statements and criticizing people.

 

It is essentially impossible to "prove" that one is not cheating, however, what I think Biker meant was that there needs to be much more independent verification.

 

The problem with that is twofold:

 

1 - Who's going to pay for it?

 

2 - Independent verification is going to require procedures, and those procedures will not be comprehensive because nothing is comprehensive, aside from the comprehensive statement that nothing is comprehensive (but enough about Russell's Paradox), therefore independent verification will not eliminate cheating because there's going to be exploitable holes in the test procedure.

 

I mean, I'm not saying it's a bad idea, but it's not going to solve the problem.

 

I agree, but if we had common standards between markets it becomes easier to do those random inspections.

[SoonerLS]

Personally I think the guy that found this was tipped off as to what to start to looking for by someone at an OEM.

NPR interviewed one of the scientists this morning. They (the scientists) initially thought they were screwing up something in their testing, so I don't think they were tipped by anyone.

http://www.npr.org/2015/09/24/443053672/how-a-little-lab-in-west-virginia-caught-volkswagens-big-cheat

[jpd80]

Ok. How much money are you all prepared to spend to catch one particular kind of cheat?

How much did those Euro researchers spend to check 35 vehicles and find anomalies?

 

These types of on-road testing real world driving tests are along the lines of what the EDE was proposing post 2017.

 

 

I agree, but if we had common standards between markets it becomes easier to do those random inspections.

Yes, As Richard said early, any test is able to be defeated but we need to look at what the standard tests really achieve beyond repeatability.

 

The issue here is how can authorities be confident that vehicles for the most part are compliant to the emission limits prescribed.

Since most driving is done on light throttle (less than 50%), we can conclude that simply driving around without goosing the throttle

is going to give a reasonable idea of the emissions profile.

 

NOX levels 40 times higher than acceptable stands out like dogs balls..

Edited September 25, 2015 by jpd80

[RichardJensen]

How much did those Euro researchers spend to check 35 vehicles and find anomalies?

 

It's generally going to be cheaper to catch cheaters the first time than it is going to be to screen for that method of cheating on an ongoing basis (think shoe bomber guy--some passenger noticed him trying to set his shoe on fire = no cost discovery of the bomb. Ten years later, think of the cumulative costs incurred in screening shoes before takeoff)

[jpd80]

Sorry All, Big Edit.......

 

Richard, I hear what you're saying about stopping all fraud in all tests,

I simply feel that more could and should be done to close some loop holes,

.

A compelling argument is mounting for gasoline and electrification, the diesel advocates have now

suffered a terminal hit - one that will be hard to recover from. Either way, diesels will now be on a short lead.

Edited September 26, 2015 by jpd80

[kyle]

Well they will have at least $7 billion to start the funding!!!

 

Does anyone know if VW has a stop order in Europe or other parts of the world? I am wondering if these vehicles continue to be sold elsewhere when they are not compliant for many markets.

[fuzzymoomoo]

Well they will have at least $7 billion to start the funding!!!

 

Does anyone know if VW has a stop order in Europe or other parts of the world? I am wondering if these vehicles continue to be sold elsewhere when they are not compliant for many markets.

I wish I could find the article again, but I read after work this morning that Switzerland has a stop sale ordered on all 1.2, 1.6 and 2.0 liter diesels.

[akirby]

I wish I could find the article again, but I read after work this morning that Switzerland has a stop sale ordered on all 1.2, 1.6 and 2.0 liter diesels.

http://www.blueovalforums.com/forums/index.php?/topic/59722-vw-caught-cheating-on-emmissions/?p=954716

[bolt in blue]

A lot of developing countries have no emission standards or safety standards and if they have them, they're not enforced at all. If WV has a hard time making these cars compliant and ends up cancelling diesel sales in the US, it wouldn't shock me if they shipped the unsold inventory to Latin America.