Well, seeing how one is titled the "Chief Engineer" I wouldn't be too surprised that he represents an entire department (or many) and is the last stop before board of directors.
To me, having been in the auto industry as a peon, Chief Engineer sounds like a person with a Junior Middle Manager pay level without officially having any direct reports.
That's odd, I work in an auto company, our chief engineer is the head of all engieering, he reports to the GM, who reports to the board of directors.
Another odd thing though, is that VW apparently puts their compliance responsibilities under engineering management. Our compliance department, which I work in, is an independent department. That has its own difficulties with being less integrated with development, but we don't have any competing goals.
"regulators are considering steps to tighten emissions standards for diesel engines"? - interesting that the old emissions standards were (one presumes) absolutely fine, until somebody turns out to have been cheating.
Well, enforcement of the old emissions standards was lacking. If "tightening" means "actually enforce the ones we have, even against the revealed gamesmanship", that's a very proper step.
I think you've got to look at this as a process of gradual refinement in which both government and industry participate. The government wants to achieve its overall objectives of reducing emissions over a long period of time (decades); it knows that if it just set the emissions targets at the target on day zero the entire industry would fail to meet them and the industry would collapse. So the regulations have to be set with some idea of what is currently achievable or within reach of the industry given an economically feasible amount of investment.
So the aim of the regulation is to push the industry in the direction you want them to go at a rate which gets everyone to the destination without destroying the industry. And every so often the regulations need to be tightened to keep the development process moving forward.
The rest of the commenters are quick to dismiss this article as presenting scapegoats, but I'm very keen to see more in-depth investigation about how this was orchestrated.
There are a lot of different ways that this type of cheating can develop over time, and simply saying "lots of people knew must have known about this" isn't very satisfying to me.
There are plenty of ways that upper management can stay in the dark. In the case of Deepwater Horizon, BP didn't make the decisions that lead to the blowout, but created excessive pressure for cost-cutting. In the case of Barings Bank, the managers were genuinely unaware of the rogue trader's fraud, but allowed him to bypass internal controls.
There are important lessons that will come out of the VW case and I'm very interested to learn them. The dismissiveness of the comments here is really disappointing, on the other hand.
Knowledge of the act, not knowledge of the law. There are very few contexts in which you can be punished for not knowing an illegal act took place, and they are all situations where you should have known but were negligent.
negligence casts a very wide remit, but applies mostly to professionals (ie, credentialed staff). general executives are granted considerable leeway with respect to negligence in many context it would appear through things like the 'business judgement rule', where bein incompetent is a legitimate excuse and not presumtive negligence. INAL but just throwing this out there for discussion to see some other sides of the issue.
There is the fiduciary duty aspect in play here. If a the employee of a business does something criminal while conduction the companies business, then it's possible that the owners are victims as well. However if they willfully created incentives and benefited from said criminal actions, it's hard to how they are not perpetrators themselves.
Try telling your accountant "I want to pay less tax and I don't care how you do it."
When you're audited, have fun claiming you didn't know that you were committing tax evasion.
Try buying something for a ridiculously low price down the pub. When the police come to take the stolen goods away, try claiming you didn't realise something selling for a tenth of its value was stolen.
In warfare. A commander can be criminally liable for failing to prevent criminal actions by their subordinates. Most famously, Tomoyuki Yamashita. Or from the prosecution brief during Ernest Medina's trial:
>'When troops commit massacres and atrocities against the civilian population of occupied territory or against prisoners of war, the responsibility may rest not only with the actual perpetrators but also with the commander. Such a responsibility arises directly when the acts in question have been committed in pursuance of an order of the commander concerned. The commander is also responsible if he has actual knowledge, or should have knowledge, through reports received by h'un or through other means, that troops or other persons subject to his control are about to commit or have committed a war crime and he fails to take the necessary and reasonable steps to insure compliance with the law of war or to punish violators thereof.'
Well, I believe that there will obviously be an effort to scapegoat some low level suckers, and, as such, the whole effort is garbage.
The basic principle is liar's reports are worthless. When reporters want to create a particular outcome (no executives should do hard time), what you end up with is fundamentally dishonest. And since we have reasonable doubts about the integrity of the investigation -- it's being performed by employees of the original deceivers -- the whole thing is garbage. You can't take it and unshade it a little bit; people who have proved their untrustworthiness simply cannot be trusted. Not just a little bit; not at all.
ps -- what constitutes a believable report? One done by criminal investigative services belonging to a country where the government doesn't own a major stake in the company being investigated.
>There are plenty of ways that upper management can stay in the dark. In the case of Deepwater Horizon, BP didn't make the decisions that lead to the blowout, but created excessive pressure for cost-cutting. In the case of Barings Bank, the managers were genuinely unaware of the rogue trader's fraud, but allowed him to bypass internal controls.
I read this as an exec saying: "Just get it done. I don't care how it's done, and I don't want to know how it's done, just do it."
May have watched too many movies, but whatever is more plausible is usually correct.
As an engineer who worked in a couple plants and focused on QC and the like, I damn near have an allergic reaction when any level of management says something like that. I immediately assumed that sort of conversation took place, and it's a strong signal something unethical is afoot - or at least a call for something unethical. ("I'm not saying you should ignore the samples, but sometimes you get different results if you test again, you know?")
>There are plenty of ways that upper management can stay in the dark.
If they knew about it, they were wrong. If the didn't know about it, they were derelict/incompetent in their duty to supervise, AKA, wrong.
> In the case of Deepwater Horizon, BP didn't make the decisions that lead to the blowout, but created excessive pressure for cost-cutting.
Oil drilling operations are very much different from any other company structure/hierarchy you are probably familiar with. BP's "Company Man" (Jobsite supervisor) made those decisions and BP gets to own the consequences. BP also made a lot of other poor decisions along the way, and afterward, which are also their responsibility. None of the parties comes out of the DWH disaster looking good, but BP gets primary responsibility, and they earned it.
I think you're approaching this from a justice angle: who is responsible, who should pay. I think akamaka is approaching this from a system perspective: what enabled this to happen, what can we change to avoid it?
It seems like the majority of people saying these are scapegoats didn't read the article that closely. It clearly says these two engineers were top aides to the former CEO.
Regulator's aren't stupid enough to be fooled by a couple of junior engineers being offered up as a sacrifice. They'll definitely go after the highest-level people they can get.
A top aide to the CEO sounds like exactly the kind of person who would be the fall guy in a scandal like this. Important enough for the regulators to be happy, but not someone who has actual power in the company.
> not someone who has actual power in the company.
Two men are named in the article. One is on the Audi board of management. The other is on the Porsche board of management.
I understand where the cynicism comes from, I really do, but it's exhausting to wade through all the uninformed naysaying in this discussion. The above information took me less than a minute to find.
>Regulator's aren't stupid enough to be fooled by a couple of junior engineers being offered up as a sacrifice. They'll definitely go after the highest-level people they can get.
Maybe in a just world. I don't hold out too much hope.
I think everybody is pissed. Just look at the message: They could not solve a technical problem, so they cheated. Every German engineer must be pissed. And everybody who has to explain, that this is not German industry standard. (at least I hope it is not)
These two are not junior engineers.
And I doubt that only two people knew about it. If Volkswagen developed diesel engines for the US market and problems arise to meet the requirements these problems have certainly not been solved in a dark back room, over a beer by two engineers. This must have been a topic on a high level. And also the solution must have been discussed.
These engineers seemed to have gained quite a reputation for their work, and were thought to be very valuable to VW. It's possible that the "lots of people" that should have known about the software simply trusted these engineers and their work to the extent that they didn't bother to inspect the software for flaws as closely as they should have.
I can't imagine something this important not having to go through multiple people for inspection. So of those people that are saying they "don't know", though it may be true, for some of them it was their job to know, I think that should (and probably is) being taken into account in the investigations.
It will be interesting to see if the investigators manage to obtain an email chain which will show where the pressure to do this came from - from day 1 I've felt it was unlikely that this was an isolated incident, or one which management weren't tacitly aware of, and I would be very surprised if it doesn't turn out that we find out where it all came from - my bet is that it was pressure from above that led to it, but that's just a hunch - it may be that management didn't know what was going on, and the engineers took it on themselves to cheat. But either way, I don't think we're near the bottom of this yet.
> in August 2007,[..] It rebranded the company’s diesel engine TDI—for turbocharged direct injection.
Huh? VW's diesel engines had been TDIs for well over a decade at that time, for example my 1996 Golf was a TDI. The TDIs are awesome engines. I previously had a gasoline engine with the same horsepower rating, but the TDI was better in every way: unkillably high torque at low RPMs (I could go from a standing start to 4th gear without ever pushing the gas pedal), great acceleration at high speeds (at 160 km/h, hit the pedal and it goes, whereas with the gasoline engine it was hit the pedal and ... wait ... wait ... still trying to accelerate ... wait ). And of course far, far better gas mileage.
To be fair, great gas mileage is true for every manufacturer. It's inherent in the technology, and applies across the full range of engine sizes and configurations. I've driven my BMW 330d 35,000 miles and to my knowledge it's never used even a single millilitre of gasoline.
The reason people perceive diesels to be better is that they have much flatter torque vs RPM curves, so any schmuck can make them accelerate without shifting to the correct gear. If diesel engines gave better performance at the same amount of horsepowers, and used less fuel, why don't we see diesel engines used in horsepower-limited racing car classes? Why has Ferrari and Lamborghini not ever offered a single diesel engine?
Audi did make a diesel race car. It was quite famous actually.
That said, the main issue with diesel is that while they have a much better torque curve, diesel engines have a much lower RPM limit. This plus the emissions issues make for pretty big compromises.
Diesels are great for having a wide spread of torque from very low RPM which makes them very tractable for everyday driving. Race cars have different priorities. They need maximum power which generally results in a very peaky engine. They substitute a highly skilled driver for tractability.
"The Audi R10 TDI, usually abbreviated to R10, is a racing car from the German car manufacturer Audi. The car is a classic at Le Mans, winning every year since its introduction until it was replaced by the R15. [..] It was the first diesel-powered car to win either of those events. "
In my opinion, this is going to get far worse for VW. People bought cars expecting the stated fuel economy - if they submit their cars to the recall, there's likely to be a class-action suit aiming to recoup enough cost to pay for the difference in fuel costs over the life of the car.
Cars that don't go through the recall process will continue to pollute at a greater rate and will earn the ire of the EPA (here in the US).
It's going to cost them a huge amount of money to satisfy both the EPA and their customers.
It's going to get worse, but not for the reason you state. Changing car performance (acceleration profile, etc.) for the worse is a necessary outcome of fixing this issue. That's why people will sue.
You're right - I was assuming the pollution controls would only result in a lower fuel economy. On the other hand, degrading the car's performance to improve emissions is the opposite end of the spectrum. Since fuel economy versus performance provides an entire spectrum of tuning options, it will be interesting to see exactly which trade-offs they make in the ECM's code.
It seems premature to assume they can fix the problem with code changes alone. If that's the case, the performance drop will likely be beyond acceptable. The required changes could make the vehicles physically unsafe to be on the road (think, being unable to accelerate out of danger's way because now you need 60 seconds to go from 0-50 km/h). If the option for a compromise between performance and regulatory requirements is already possible with the current hardware, wouldn't the cars likely have already shipped as such?
It seems much more likely that there will be parts replacement involved, and we may soon discover that the entire engine must be swapped out. This is unlikely to be feasible for cost, but bricking cars to intolerable driving conditions is also not enough. It's going to be really interesting to see how this develops. Done properly, this would probably be the end of VW. So I'm sure we'll see some half-assed solution that doesn't work for the consumer, but lets the company survive for a few more years until they finally collapse from consumers taking their future business elsewhere.
You want to throw Barney "Roll the Dice" Frank in jail?
There tend to be lots of unintended consequences to punishing stupid politicians like that, or people outside the government following the government's prior orders.
"For more than a year, Volkswagen executives told the Environmental Protection Agency that discrepancies between the formal air-quality tests on its diesel cars and the much higher pollution levels out on the road were the result of technical issues, not a deliberate attempt to deceive Washington officials."
So VW is telling us that this year of lying to the EPA was done completely by these two engineering managers and nobody else in the VW executive staff?
Where did they say that? They are suspending people left and right and there are investigations ongoing against Winterkorn.nright now there has nothing final been said yet.
glad they found the culprits. I was absolutely certain it would be renegade engineers and the higher level up (politically secured allies) have nothing to do with it. Torches and pitchforks everyone.
The amount of restraint I have to exercise as an ex-consultant who made his living working for the likes of Audi, VW & Siemens nobody here can possibly comprehend.
Even if their superiors did honestly not know that they were loading fraud-ware in the vehicles, this is most likely the result of systemic social dynamics and perverse incentives. Mutually exclusive goals get set with no regard to the laws of physics or reason. Everyone who dares mention this self selects outside of the management hierarchy, so the whole chain of command becomes an echo chamber of overconfidence.
Then, at the end of the day, something has to give. It was the emission control testing in this case. It could have been much worse.
I had an interesting conversation with a co-worker today, about why unit-test coverage is always bad unless enforced.
I decided that it's because even though my boss and myself will have good intentions, we can both mutually decide (without ever discussing it) that the tests should slide - I will try to meet his deadline, he won't check too hard.
Unless there are credible institutions that take that decision away from us (a build bot that refuses to deploy untested code for example) we will slip towards worse.
Management does not need to apply pressure towards misbehaviour - they simply have to not constantly strive for best behaviour. Our expectations fill in the rest.
As such, we are unlikely ever to find an email from the board saying "Hans, frig the tests next week, we can't be arsed to invest in catalytic converters", but the organisation decided that anyway.
I think it's to do with lack of candour - see the Pixar book out recently
This is not on the same league. Your unit-test coverage example is closer to the Toyota incident. Surely, it was a catastrophe, and lives were indeed lost, but no one actively inserted a "speed up when driver hits the breaks" feature into the firmware.
A closer analogy to the VW case is if the build bot actually did exist, but you wrote some nefarious unit test that just spit out all OK results regardless of what the exercised code did, so that you were allowed to merge whatever crap your team produce, as long as it actually compile.
If that is the case, we can discuss different possible assignations of responsibility, but you cannot claim that some good intentioned people just slide into misbehavior due to not paying enough attention. Someone had to be aware they were actively cheating!
These scapegoats are old dudes; they might really be the right scapegoats (as opposed to the a couple of junior or mid-level coders, one might expect).
Do I miss that in the article, or is it really not stated from where the information that the investigation is focusing on them is coming? People "familiar with the matter" as only source? It would be a pretty bold (and irresponsible) move from a criminal investigation to leak suspects to the press that early, given that this will destroy their career. Stuff likes that needs prove or very very strong hints pointing at them, enough to start a trial, and there is no word about that.
I understand everyone who thinks that this looks like presenting scapegoats. This stinks. The whole article shouldn't have been published like that, and we should not upvote it.
> Both men developed an international reputation as engineers, said Ferdinand Dudenhöffer head of the Center of Automotive Research at the University of Duisburg-Essen.
> “It will be difficult for Volkswagen to replace them,” he said.
Genuinely legendary engineers, and allegedly with the balls to cheat (but not quite the balls to stand on principle). It will indeed be hard to find that combination among the remaining legends.
Why didn't this same link I posted earlier get picked up when peterthomas submitted it? I think they are the same? Seems like HN had a glitch? https://news.ycombinator.com/item?id=10330573
It is outrageous to suggest that the blame lies with a handful of people and not a whole division of the company. I don't work in the automotive industry, but I do work as a design engineer in the semiconductor industry and for a mature product in a highly competitive industry, I can guarantee that the external test wasn't the only test that these cars were put though. When we release a chip external reviewers will run it thorough a handful of benchmarks, but we run far more tests internally because when you are building complex systems, not everything that you plan for happens, and you need an extreme level of detail to be able to understand and fix performance problems. Literally hundreds of engineers will participate in performing these measurements, and most of them, along with the whole management tree up to the CEO (especially for an important consumer facing product), will participate in interpreting them. Not doing this means getting left in the dust by your competitors. There is no way that only a small number of people knew about this. Two is laughable, people here with software engineering backgrounds should be highly skeptical that any piece of software in a product with as many users as a car was only examined by two people. The only explanation is that a large number of engineers and managers knew about this and did nothing.
One of the people named is Audi’s chief engineer, and they're both described as "being in charge of R&D" at VW. So I would presume they're pretty high up the food chain, not the low-level guys who actually wrote the if: test then: cheat bits.
Still I can't believe that so many people must have known about this and just let it happen. Sad. This is a tech oriented forum and I hope some of the engineers reading this would act differently in a similar situation.
Strangely, nobody talks about the investors... Did they command such a cheat ? After all, the top engineers are already 2 levels deep in the management (top level being investor, then CEO, then tehe engineers) so they're relatively low enough to be scapegoat...
99 comments
[ 3.3 ms ] story [ 93.3 ms ] thread"head of" is a low level (entry-level) mgmt position in these places. No idea why anyone would down-vote you.
Another odd thing though, is that VW apparently puts their compliance responsibilities under engineering management. Our compliance department, which I work in, is an independent department. That has its own difficulties with being less integrated with development, but we don't have any competing goals.
So the aim of the regulation is to push the industry in the direction you want them to go at a rate which gets everyone to the destination without destroying the industry. And every so often the regulations need to be tightened to keep the development process moving forward.
There are a lot of different ways that this type of cheating can develop over time, and simply saying "lots of people knew must have known about this" isn't very satisfying to me.
There are plenty of ways that upper management can stay in the dark. In the case of Deepwater Horizon, BP didn't make the decisions that lead to the blowout, but created excessive pressure for cost-cutting. In the case of Barings Bank, the managers were genuinely unaware of the rogue trader's fraud, but allowed him to bypass internal controls.
There are important lessons that will come out of the VW case and I'm very interested to learn them. The dismissiveness of the comments here is really disappointing, on the other hand.
Knowledge of the act, not knowledge of the law. There are very few contexts in which you can be punished for not knowing an illegal act took place, and they are all situations where you should have known but were negligent.
When you're audited, have fun claiming you didn't know that you were committing tax evasion.
Try buying something for a ridiculously low price down the pub. When the police come to take the stolen goods away, try claiming you didn't realise something selling for a tenth of its value was stolen.
>'When troops commit massacres and atrocities against the civilian population of occupied territory or against prisoners of war, the responsibility may rest not only with the actual perpetrators but also with the commander. Such a responsibility arises directly when the acts in question have been committed in pursuance of an order of the commander concerned. The commander is also responsible if he has actual knowledge, or should have knowledge, through reports received by h'un or through other means, that troops or other persons subject to his control are about to commit or have committed a war crime and he fails to take the necessary and reasonable steps to insure compliance with the law of war or to punish violators thereof.'
The basic principle is liar's reports are worthless. When reporters want to create a particular outcome (no executives should do hard time), what you end up with is fundamentally dishonest. And since we have reasonable doubts about the integrity of the investigation -- it's being performed by employees of the original deceivers -- the whole thing is garbage. You can't take it and unshade it a little bit; people who have proved their untrustworthiness simply cannot be trusted. Not just a little bit; not at all.
ps -- what constitutes a believable report? One done by criminal investigative services belonging to a country where the government doesn't own a major stake in the company being investigated.
I read this as an exec saying: "Just get it done. I don't care how it's done, and I don't want to know how it's done, just do it."
May have watched too many movies, but whatever is more plausible is usually correct.
If they knew about it, they were wrong. If the didn't know about it, they were derelict/incompetent in their duty to supervise, AKA, wrong.
> In the case of Deepwater Horizon, BP didn't make the decisions that lead to the blowout, but created excessive pressure for cost-cutting.
Oil drilling operations are very much different from any other company structure/hierarchy you are probably familiar with. BP's "Company Man" (Jobsite supervisor) made those decisions and BP gets to own the consequences. BP also made a lot of other poor decisions along the way, and afterward, which are also their responsibility. None of the parties comes out of the DWH disaster looking good, but BP gets primary responsibility, and they earned it.
Regulator's aren't stupid enough to be fooled by a couple of junior engineers being offered up as a sacrifice. They'll definitely go after the highest-level people they can get.
Two men are named in the article. One is on the Audi board of management. The other is on the Porsche board of management.
I understand where the cynicism comes from, I really do, but it's exhausting to wade through all the uninformed naysaying in this discussion. The above information took me less than a minute to find.
Maybe in a just world. I don't hold out too much hope.
I can't imagine something this important not having to go through multiple people for inspection. So of those people that are saying they "don't know", though it may be true, for some of them it was their job to know, I think that should (and probably is) being taken into account in the investigations.
Huh? VW's diesel engines had been TDIs for well over a decade at that time, for example my 1996 Golf was a TDI. The TDIs are awesome engines. I previously had a gasoline engine with the same horsepower rating, but the TDI was better in every way: unkillably high torque at low RPMs (I could go from a standing start to 4th gear without ever pushing the gas pedal), great acceleration at high speeds (at 160 km/h, hit the pedal and it goes, whereas with the gasoline engine it was hit the pedal and ... wait ... wait ... still trying to accelerate ... wait ). And of course far, far better gas mileage.
That said, the main issue with diesel is that while they have a much better torque curve, diesel engines have a much lower RPM limit. This plus the emissions issues make for pretty big compromises.
https://en.wikipedia.org/wiki/Audi_R10_TDI
Cars that don't go through the recall process will continue to pollute at a greater rate and will earn the ire of the EPA (here in the US).
It's going to cost them a huge amount of money to satisfy both the EPA and their customers.
It seems much more likely that there will be parts replacement involved, and we may soon discover that the entire engine must be swapped out. This is unlikely to be feasible for cost, but bricking cars to intolerable driving conditions is also not enough. It's going to be really interesting to see how this develops. Done properly, this would probably be the end of VW. So I'm sure we'll see some half-assed solution that doesn't work for the consumer, but lets the company survive for a few more years until they finally collapse from consumers taking their future business elsewhere.
I hear people say the general statement a lot, but I'm not sure who specifically they're talking about.
There tend to be lots of unintended consequences to punishing stupid politicians like that, or people outside the government following the government's prior orders.
http://www.houstonchronicle.com/business/article/Volkswagen-...
So VW is telling us that this year of lying to the EPA was done completely by these two engineering managers and nobody else in the VW executive staff?
The amount of restraint I have to exercise as an ex-consultant who made his living working for the likes of Audi, VW & Siemens nobody here can possibly comprehend.
e: And hilariously for you, according to Wikipedia, Piëch himself is in fact a dyslexic atheist. Zounds!
Even if their superiors did honestly not know that they were loading fraud-ware in the vehicles, this is most likely the result of systemic social dynamics and perverse incentives. Mutually exclusive goals get set with no regard to the laws of physics or reason. Everyone who dares mention this self selects outside of the management hierarchy, so the whole chain of command becomes an echo chamber of overconfidence.
Then, at the end of the day, something has to give. It was the emission control testing in this case. It could have been much worse.
I decided that it's because even though my boss and myself will have good intentions, we can both mutually decide (without ever discussing it) that the tests should slide - I will try to meet his deadline, he won't check too hard.
Unless there are credible institutions that take that decision away from us (a build bot that refuses to deploy untested code for example) we will slip towards worse.
Management does not need to apply pressure towards misbehaviour - they simply have to not constantly strive for best behaviour. Our expectations fill in the rest.
As such, we are unlikely ever to find an email from the board saying "Hans, frig the tests next week, we can't be arsed to invest in catalytic converters", but the organisation decided that anyway.
I think it's to do with lack of candour - see the Pixar book out recently
A closer analogy to the VW case is if the build bot actually did exist, but you wrote some nefarious unit test that just spit out all OK results regardless of what the exercised code did, so that you were allowed to merge whatever crap your team produce, as long as it actually compile.
If that is the case, we can discuss different possible assignations of responsibility, but you cannot claim that some good intentioned people just slide into misbehavior due to not paying enough attention. Someone had to be aware they were actively cheating!
I understand everyone who thinks that this looks like presenting scapegoats. This stinks. The whole article shouldn't have been published like that, and we should not upvote it.
> “It will be difficult for Volkswagen to replace them,” he said.
Genuinely legendary engineers, and allegedly with the balls to cheat (but not quite the balls to stand on principle). It will indeed be hard to find that combination among the remaining legends.