Hacker News new | comments | show | ask | jobs | submit login
Nissan Admits Internal Emissions-Test Results were Falsified (wsj.com)
665 points by blawson 37 days ago | hide | past | web | favorite | 342 comments

Non-Paywall archived version:


This isn't what you think it is from the headline.

This is not a case of Nissan software engineers writing emissions test defeat code. Rather, this is some Nissan factory workers faking internal emissions test results that are done on every completed car. Only about 1,000 cars were affected, and almost all of these cars still met regulatory emissions limits - the targets they were trying to meet were Nissan's stricter internal standards.

> Other than one model still under investigation, Nissan said all of the affected vehicles met Japanese auto standards. The data alterations were made by staff to meet Nissan’s own stricter internal standards, it said.

Good catch. This comment should probably be higher, because I don't think the majority is reading the news that way.

> I don't think the majority is reading the news that way

More like, "I don't think the headlines are written to portray the news that way"

I put the blame, always, on sensationalist clickbait headlines from dishonest news organizations. They are competing for your attention and clicks, so the headline is often flat out wrong.

Quite, and the first two paragraphs of the article, which are all I can read because I'm not a WSJ subscriber - which I suspect will be the case for many here - do nothing to dispel the sensationalist interpretation.

(As an aside I'm starting to opine that posting WSJ articles here may be bad form for this very reason.)

I've only been lurking a short while, but Ive been finding WSJ to be relatively unreliable.

What's the sensationalism? Nissan was producing out-of-spec cars due to falsified data, no?

When I read the headline, it seemed to me that Nissan was falsifying data to meet Government/mandated emissions tests and was caught.

In reality, Nissan's internal workers were caught falsifying an internal test. That might seem like a small point but it's quite an important difference - the difference between a live malicious data breach and an internal audit failing.

Nissan's specs are more strict than whatever Government test, so the Government test is not really relevant.

The bigger point is that Nissan would be committing fraud on its customers by quoting false specs for the products they are selling.

I just read the article and didn't see where it said the internal standards were quoted to customers. Could you point to that part?

I never said it did. Specs are usually pasted to a car's window.

One of those is a crime, and one of those is an internal problem. One of those results in extensive fines and jail, one results in finger wagging from the authorities.

Either way it is a crime in most jurisdictions.

For example, in DC there is fraud in the first degree and fraud in the second degree, as well as false promise to future performance: https://code.dccouncil.us/dc/council/code/sections/22-3221.h...

>(a) Fraud in the first degree. — A person commits the offense of fraud in the first degree if that person engages in a scheme or systematic course of conduct with intent to defraud or to obtain property of another by means of a false or fraudulent pretense, representation, or promise and thereby obtains property of another or causes another to lose property.

>(b) Fraud in the second degree. — A person commits the offense of fraud in the second degree if that person engages in a scheme or systematic course of conduct with intent to defraud or to obtain property of another by means of a false or fraudulent pretense, representation, or promise.

>(c) False promise as to future performance. — Fraud may be committed by means of false promise as to future performance which the accused does not intend to perform or knows will not be performed. An intent or knowledge shall not be established by the fact alone that one such promise was not performed.

Its not a bigger point. If it was, they wouldn't have written the headline to deliberately mislead away from that fact.

The specifics of the test don't really matter regarding selling a vehicle with fraudulent specs.

Of course it does. The bare minimum to be "safe" is much different than strict internal standards. All rules are not created equally, and breaking different rules should be treated differently. This is why the purposeful misdirection in the headline matters - it is not true or accurate.

How was the headline not accurate?

Because outspoken HN readers assumed this was something to do with government?

> How was the headline not accurate?

Because it does not provide it's readers with an understanding of what happened, purposefully, to create a more sensationalist headline to generate more clicks and likely more revenue through advertising and subscriptions.

> Because outspoken HN readers assumed this was something to do with government?

First, this is needlessly insulting. Second, it is common knowledge that emissions standards are government regulations (because they are), so titling an article about a car manufacturer about failing emissions standards points _almost everyone_ in the wrong direction.

Were you in the newsroom? How do you know the purpose was to "create a more sensationalist headline" (which wasn't sensationalist).

>First, this is needlessly insulting. Second, it is common knowledge that emissions standards are government regulations (because they are), so titling an article about a car manufacturer about failing emissions standards points _almost everyone_ in the wrong direction.

As insulting as assuming the motivations of an author? I don't think so. Emissions standards are also builder specs. What came first: builder specs, or government specs?

>titling an article about a car manufacturer about failing emissions standards points _almost everyone_ in the wrong direction.

I disagree. The title was factually accurate. The current title is arguably inaccurate now, as I doubt Nissan even used the word 'internal' in their 'admission'.

It's a lie by omission. The title should have been something like: "Nissan Admits Internal Test Data Was Falsified"

Boring, I know.

Lie by emission.

And who has verified what Nissan said?

Ok, we'll put "internal" in the title above. Thanks!

Hey man, it's still not clear. I would say "Nissan Admits Data for Internal Emissions-Test Was Falsified", because right now it seems like the data was internal, not the test.

I second this, I came after the change and I still didn't find it clear at all.

Nissan externally admits internal data for internal test was internally falsified.

How about "Nissan Admits Internal Emissions-Test Results were Falsified"

Ok, we'll go with that. I'm not super convinced we've improved things but you guys win.

I'm not saying change it again, but that sounds like the emissions are occurring internally.

So on an average day for you, how much time do you spend dealing with whiny posters complaining about titles, and then manually re-editorializing them yourself?

I see you've had to respond to this one at least twice, and that doesn't seem unusual from what I see here. Multiply it by a couple dozen articles that are on/near the front page per day, and...

You should keep a log and see, I suspect it's a larger fraction than you realize. It's not a positive benefit for the community discussion either.

HN users' passion for accurate titles is an unstoppable force, so the object it meets had better be moveable. I learned that the hard way. Title fever is not good for discussion, that's true; but the only way to minimize it is to flex with objections. I learned that the hard way too.

The interesting question is why. It can't just be about titles; those are too trivial to bring out that kind of emotion.

Context switching sucks is why.

An accurate headline is like parking in a parking spot. A misleading clickbait headline is like backing into the same parking spot.

This is the only place where we have any control over that, every other social media venue we're stuck with the flashing lights and beeping noises.

Thanks for your vigilance.

Thank you so much for pointing this out!

Whoa. Pretty major difference ther.

Thanks for saving me a pointless read

As software engineers/computer people, we have to have ethical standards. If you participate in writing code like the one in the VW case then you are not only complacent but active in fraud.

Engineers have a code of ethics. If you were to knowingly build an unsafe bridge then you would be liable for the damage and injuries it caused if/when it failed. This is true even if your boss told you to do it. You not only have a moral/ethical obligation to not build that bridge, but also a legal one.

As we write software we need to keep these things also in mind. We are writing things that affect peoples lives. That could kill people. That affect the environment. We need to hold ourselves to the same moral obligations as other engineers. We have a duty to prevent harming people when we write our software.

If you have no stand to refuse to do the work then it is your duty to report such actions. First up the chain, and if that is rejected, then to the media. It is also your duty to report it if you have refused and you see it still happening. There are whistleblower rights in place to protect you.

Software engineers don't understand they're not professionals. I mean, they describe and think of themselves as such, but they are not legally such.

Being a "professional" legally means that you're given exclusive rights to perform some economic activity and in return for this (usually) lucrative right, you have a great number of obligations - being responsible for your work (ie, facing the possibility of never being able to perform said economic activity legally again), an enforceable code of ethics, an dispute / resolution process etc, steep fees, continuing education, an accreditation exam, etc.

Software engineers by and large do not want the obligations and restrictions of being a real profession, so no, they don't have any special ethics or rules to report or do anything in these cases beyond what a designer or secretary would do.

I've never heard of this legal definition of "professional" that you put forward. Where is this in state or federal law?

In law school, we were taught that a professional has obligations that go beyond his or her own self interest. Unlike a real estate developer or a widget manufacturer, an attorney has an obligation to do what's right for the client even if that causes the attorney to lose money or imposes a competitive advantage. Attorneys, for example, can't engage in the kinds of advertising speech that would be entirely lawful if done by a car dealership or furniture store.

This traditional definition is hundreds of years older than the bar exam or ABA code of ethics, and older than formal graduate law school (a 20th Century American invention). All of that came later, after lawyers understood themselves differently.

I would encourage people who write code -- software engineers, developers, coders, hackers, however one chooses to describe this work -- to think of themselves as professionals with obligations that go beyond their own self interest.

I would imagine OP is referring to "Professional Engineer".


Every single design, bridge, etc is certified by a Professional Engineer. Your ass is on the line. I don't mean just being fired from your job - but jail time.

Per the website:

"A century ago, anyone could work as an engineer without proof of competency. In order to protect the public health, safety, and welfare, the first engineering licensure law was enacted in 1907 in Wyoming. Now every state regulates the practice of engineering to ensure public safety by granting only Professional Engineers (PEs) the authority to sign and seal engineering plans and offer their services to the public. "

I suspect, and hope, software used critically have the same. Yes it's more bureaucracy and overhead...but being in software, I know any joe schmoe can walk on and develop code without oversight.

I'm not saying if you're making the next twitter, that you need a PE. Or that every developer needs to have a PE license - but like other industries there is one PE per X amount of non-PE engineers.

Due to the Threac-25 incident (https://en.wikipedia.org/wiki/Therac-25) there are standards for medical software (https://en.wikipedia.org/wiki/IEC_62304).

I'm not in medical so I don't know how widely adopted, or how strict it is adhered to. But it might be IEC 62304 needs to be branched out.

In the US at least, the regulation of software in medical devices regulates the process by which the software is developed and tested.

The qualifications of the people developing the code are somewhere down in the process-stack; there are no specific requirements [1].

In general, to follow "Current Good Manufacturing Practices" (a term of art in the FDA design controls regulation), qualifications for the people on the team have to be documented and a paper trail kept up-to-date, to be able to answer questions of the nature "was person <X> qualified to be working in this role? How do you know?" with objective evidence.

It's up to the judgment of the engineering management of the project what they state the requirements of the role to be. If you say that it takes a high-school diploma or better to be able to develop life-safety-critical medical device software, you may pass a process audit since you document and track that requirement, but if there is an "adverse incident" I would think whoever signed off on that decision might be exposed to liability.

[1] My specific knowledge about this is ~5 years out of date so this may have changed, but I doubt it

> I would imagine OP is referring to "Professional Engineer".

In part yes in part no. PE's will be held to a MUCH higher legal standard but every engineer is taught to hold themselves to a high ethical standard. That you are not supposed to do something dangerous even if you think you won't get in trouble for it. Even if it isn't directly your responsibility you are supposed to report something that looks dangerous. PE's do the final sign off and their ass is on the line much more than the engineers that actually did all the building and designing. But the code of ethics is supposed to be there all throughout the chain.

So what I'm saying is that in software we are part of that chain. And therefore the ethical obligations should still hold. They should hold at every part of that chain. At no point in the process should anyone say "well it is someone else's ethical obligation or concern". If you touch something, if you see something, then it is part of your ethical obligation. At least as an engineer, but I would argue that it is also if you write software.

PEs exist for many branches of engineering. Including software. But the PE for software engineers is apparently being discontinued due to lack of interest.

Yes, NCEES (the body that administers the Engineering licensing exams nationwide in the US) is discontinuing their Software Engineering exam next year because only ~80 people have taken it since it was created in 2013.

I think the biggest problem they had was that despite having the exam, not many states actually issue Professional Engineer licenses in the Software Engineering discipline. Because of this, I don't see why anyone (in those states) would bother taking the exam if they can't actually get a license and use the title "Professional Engineer".

I forgot licenses were by state. It's probably also the case that software engineers don't normally have to "sign off" on things for regulators in a formal way as is the case in other areas of engineering. So, other than having a title to stick on your business card, taking the PE exam doesn't buy you a whole lot whereas it's virtually required for other engineering branches.

Well, "practicing medicine without a license" is a crime. I mean, you can call yourself a "healer" or something and claim to help people in some general, non-specific way, but you can't prescribe meds, cut people up, etc etc.

So if you're a doctor you're given exclusive right to perform the economic activity of doctoring. In return you have a steep set of obligations, which include being truthful, ethical, not refusing service, etc.

What you said is a part of these "obligations", which definitely put you at an economic disadvantage, but that's ok, because you have exclusive rights to that activity. A doctor's obligations do impose costs on them, but not compared to other doctors.

The critical bit I'm trying get at is enforceability. If a doctor acts unethically, they can be challenged in front of their board by a patient and if they lose their license they can never practice medicine again. Having your livelihood at stake is a great incentive to act properly.

A software engineer such as myself, OTOH, has no little incentive beyond my own personal ethics. In many cases, being ethical will place you at an economic disadvantage compared to other software engineers. Actually, the whole notion of "ethics" is hand-wavy and ill-defined, because there's no professional organization.

Anyway, we can wish for good things and that'll party work and it's better than being cynical, but it'll only get you so far.

> I've never heard of this legal definition of "professional" that you put forward. Where is this in state or federal law?

Professional, as in professional engineer. Think civil engineers, for example, who need to be registered with the board of engineers in order to provide engineering services (eg. approving plans for construction, supervising construction).

Other countries also have specific legal statutes that regulate the title of engineer, such as Canada.


1. A person, who is a member of a professional body due to the education qualification and follows the prescribed moeal and prefessional code of conduct.

2. A person who has mastered a high level of expertise in a subject, notion on field.

Completely different definition. Dawn can refer to both a time of day and dish soap, but they are not interchangeable definitions.

> PROFESSIONAL MISCONDUCT 'Behavior different from what is prescribed in the moral, ethical and professional code of conduct'

For example only uses definition 1.

But when an individual says that they are a "professional", they could easily be using either definition, and still be using the word correctly. The OP said "Software engineers don't understand they're not professionals..." as if only the first definition was correct.

There's an incentive for lawyers, doctors, and accountants. You get a license to work from a board of peers, there's peer review, you can lose the confidence of that board and have the license to work revoked. I don't know the histories of those professions, and what's in common other than "obligations that go beyond self-interest". But I think lawyers are substantially different from other professions, even setting aside the state bar: they are officers of the court. They're an integral part of the legal system and they have a superseding duty to that system's function, not only to a client.

So how would software developers/programmers/etc create something equivalent to that? Seems to me some critical mass of that industry would need to lobby state and/or federal government for the creation of a board who then produces a standard and certifies members conform to the standard under penalty that if they do not police their own members that the state can revoke the franchise/monopoly on that certification.

“Attorneys, for example, can't engage in the kinds of advertising speech that would be entirely lawful if done by a car dealership or furniture store.”

Why does Saul Goodman come to mind.

> Attorneys, for example, can't engage in the kinds of advertising speech that would be entirely lawful if done by a car dealership or furniture store.

Are you sure about that? Because I've seen some really sleazy advertising by lawyers. Personal injury lawyers in general, but Jim Adler, the Texas Hammer especially comes to mind.

Just as a quick example, the Florida Bar had a regulation that an attorney’s website could feature a picture of the lawyer, but the background was required to be either a single or gradient color or a generic courthouse.

GP didn't say that law is not a profession, but rather that computer programming is not a profession. This is obviously the case, since there are neither educational requirements nor state-approved certifications for this work. There are ethical expectations, only because all humans have ethical expectations.

IANAL but one could certainly imagine 1A challenges to any attempt to enlist the state in the enforcement of a cartel like that of the lawyers or the physicians. In that sense it's no more likely that programming would ever become a profession than e.g. that writing would become a profession.

The word "professional" is overloaded. Perhaps a better term is "government certified professional". Medical doctors and (physical) engineers must meet a prescribed set of minimum standards and take an oath.

Attempts to form a set of software standards to be used in a similar way has proven elusive. Software is too malleable, and languages and "state of the art" architectures come and go like clothing styles (often without merit in my opinion). Physics and the human body don't change very often. Software approaches do.

I prefer accredited or registered.

Legally, being a professional only means you're exempt from the requirement for overtime pay.

Someone earlier in this thread is confusing "professional" and "engineer".

I think this is really critical part. In my country a lot of (private) banks promoted the title of clerks to 'officers'. Not the title employees asked, but being officers they are not entitled to overtime.

Agreed. There’s definitely a real technical meaning in the word professional. Although the idea of being a “profession” sounds good, it’s riddled with regulation and ways that make the job less likeable for the practitioners. I left one before becoming a software developer.

Nonetheless, being an ethical human being is just a good idea.


Search for, "Member grade". The only path for a large portion of all people being called "software engineers" would not even be considered for 6 years.

Those 6 years would probably be the most formative in educating them of their ethics and responsibilities. A service I am not sure IEEE really prioritizes anyways.

IEEE can do whatever it wants. They don't certify anyone as a professional engineer. That is NSPE's job in the US.

Yeah, I am apparently not qualified to receive anything based on experience. https://www.nspe.org/membership/types-membership

Or if I want to pay them to get a certification no one cares about, I can. https://www.nspe.org/resources/education/exam-review-and-pre...

The system in that regard is broken.

They're professional. They're not engineers, in the (legally) strict sense of the word. I think that's the distinction you meant to make.

"Engineer" as a title carries significant connotations of professionalism and integrity.

If computer programmers or other developers want to refer to themselves as engineers, then it behooves them to also step up their game in the ethical field.

Too often people want the "engineer" title because of its social cachet but don't strive to leave their mediocrity behind.

Are you implying that most software engineers aren’t ethical? That isn’t true in my experience.

I wholly object to this line of reasoning. I don't see how this wouldn't blow-back on software developers and make them an easier scapegoat.

Take these car manufacturers gaming the emissions test, for example. I find it very easy to imagine management making a series of requests for changes to the software that seem innocent but allow for this exact outcome: changing the way the system performs when being tested for emissions. They may have asked the development team to add some kind of a switch so that they could toggle the behavior on and off in order to assess how overall performance was impacted.

That sounds innocent enough to me that I wouldn't be eager to risk my job (and perhaps the ability to get work in the future) over suspicions. Certainly it's not enough that I'd try to get law enforcement involved.

This idea that software developers (or engineers, if you prefer that term) should be responsible for the entire system, even though they likely only work on one component, sounds to my ear like nothing more than setting those developers up to be easy scapegoats. If you wanted to get me on board in any way, I'd need to see some system where all code is signed and, perhaps, stored in some tamper-proof write-only storage system managed by a third party. I'd need to be able to verify beyond any doubt that code a particular developer wrote hasn't been tampered with before I'd be willing to send that developer to prison, or at the least, ruin their career.

Management is at fault for requesting the feature that enabled them to game the system. Management should pay the cost, I see no reason to cast any blame on developers. I have yet to see any compelling evidence that the developers knowingly added code specifically to game the system. If evidence of that is out there, I would be super interested in reading it and would appreciate any links anyone might have.

> I find it very easy to imagine management making a series of requests for changes to the software that seem innocent but allow for this exact outcome: changing the way the system performs when being tested for emissions.

Off the top of my head, I've written many pieces of code that 'game the system' but are used only for testing purposes. I know full well that if those pieces get into production they could be hazardous, but there is nothing ethically wrong with me creating them. If a superior takes that code and uses it, that's not my fault.

It would be supremely easy for management to put in a bunch of requirements for the VW emissions system that are blatantly dangerous, but 'for testing purposes'. The developers need to be able to trust the higher-ups and certainly shouldn't be held responsible for their decisions.

To clarify on my OP, I want to say that there is a distinction between a legal and ethical obligation.

In engineering only a PE really has the legal obligation. Their on the hook if something goes wrong. But at the same time everyone involved has an ethical obligation to ensure that best practices are being followed and nothing is being done that would result in harm. This is even if they wouldn't get in any legal trouble.

Or that a doctor has both legal and ethical obligations. Their ethical obligations can many times supersede their legal ones.

I am not arguing that we who write software need to be held legally accountable, but that we need to think about any potential ethical concerns with what we write. And if you are saying "well yeah" then we're in agreement. And I think this should be pervasive. I think we should frequently talk about it.

But we are in high demand right now. Many of us have the leverage to say "I don't agree with this" and go work somewhere else. If you personally have that ability I think it should be used as well. But I understand not everyone has that leverage.

I think I understand what you're saying and I agree that no software developer should willingly write code that will be used to break or circumvent the law. I believe that if I was put in such a position that I would walk rather than implement the feature. I don't think we are alone in that feeling, I believe many developers feel the same way.

I worry that a feature meant to circumvent or break the law will not be easy to spot. Many projects are broken into smaller pieces that interact, many times the laws or rules change from state to state. Even in regular day to day work I'll see a feature request that effectively undoes another request. Often developers are working in areas in which they are not subject experts.

I think it's easy to write code that could (or even will) be misused; it may not look troubling until it's interacting with the larger system.

But that's why I'm arguing that we have to think about it. As well as why I'm not arguing for legal accountability.

We have to be aware that the things we create aren't always going to be used for what we intend them to be used for. So you frequently have to ask yourself "how could this be abused?" That is something that engineers frequently ask themselves. I don't understand why developers can't be asked to do the same.

You are totally right, I honestly agree with you. It's an important question, it's worth asking and we should all be asking it. I think there's a lot more grey in software development, grey people may not have as much of in other fields. But that does not mean we can simply ignore the question and it certainly doesn't diminish it's importance. More than that, it doesn't diminish our responsibility.

My concern really revolves around the scapegoating and protecting developers from situations where there is more grey than black or white. But that is not a sheild from personal responsibility or moral obligation, the questions you raise are important in their own right.

I definitely think the ethical concerns are a little more abstracted with software than general engineering. I mean it is more obvious how a bridge affects people than a social network.

But because of that I'm not calling for legal obligation. But I'm calling for more people thinking about the abstraction and how what they write could be abused. But it is no easy matter. Ethics are hard.

You are right. From what I have seen that most of the time even quite senior engineers are not aware of larger implications beyond technical architecture of software they are developing.

Most engineering schools have classes about engineering ethics. When I was an undergrad in UC Berkeley, we not only had an ethics class (called "Social Implications of Computing") but also every single class had ethics sections, e.g. if you're working in a mission critical OS don't forget what happened to Therac-25, if you're writing ML Algorithms don't forget what happened to Uber etc... These were actual parts of classes, mentioned by professors. I think ethics is a very important part of both engineering and being a human being. You cannot make bridges and not be responsible if they collapse. You cannot make an OS and not be responsible if a bug kills bunch of people. Similarly, you cannot write fraud software and go by your day because that's what your boss ordered you to do. Engineers are highly skilled workers and we must have enough consciousness to deny doing a job if it's not ethical. Otherwise, it's just banality of evil.

>You cannot make bridges and not be responsible if they collapse. You cannot make an OS and not be responsible if a bug kills bunch of people.

This is a bogus comparison though. A bridge has one obvious use and can't be picked up and re-used in many other bridges. An OS (particularly an open source one) has nearly unlimited uses beyond the realm of the author's control.

Let's not put Linus in prison when a kernel panic results in a Tesla slamming into a wall.

That's not what it being argued here. Not by me and I do not believe by the parent to your comment.

One of my all-time favorites: https://en.wikipedia.org/wiki/Ariane_5#Notable_launches or how a bug can blow up a rocket ;-)

I like this. At University of Florida we had 1 class. It was...alright, better than none. Everyone graduating through the college of engineering had to take it. But to have it mentioned in every single class and the relevance would be better!

this is very encouraging to read, thanks for this comment

I had a similarly titled class at a no-name state school. I think the professor mentioned something about ACM recommending/requiring the course, so it should be pretty widespread.

On a more discouraging note, there was normally single digit attendance out of triple digit enrollment in the particular course I took. I'd say average engagement was ~6 hours total over the course of the semester... Including a movie. I don't know anyone that didn't get an A. There were a few from that class that went onto big names, and they weren't among the regular attendees.

I think some sort of ethical component to the degree is required for ABET accreditation. My CE degree was retroactively accredited so I'm not sure if current students have to take a separate class, but one of the changes that was made while I was a student and the ABET process was happening was that we spent some amount of time in our project courses' lecture time (each semester has a 5-credit project course) covering ethical issues. ACM/IEEE membership (which itself, or at least "membership/sponsorship of a professional society by the school", might have been part of the accreditation requirements too since we all joined IEEE) may impose additional requirements like discussion on ethics.

I think the general point stands though that for everyone calling for ethics classes -- a lot of CS and CS related degrees like CE already have them. Furthermore many companies already have various code of conduct types of annual training. The solution has to be something else than "education".

The other thing I'd add to this is if you look at the VW and other scandals, you'll see "rogue engineers" are the first ones executives will try to throw under the bus. Ethics aside, you're taking a huge risk in order to inflate someone else's bonus.

In the end if things go sideways the company won't have your back. That boss who told you to cheat (verbally of course) will suddenly have "no recollection" of the conversation.

Yes, I don't think this is the way it actually played out. As you mention, why would such a "rogue engineer" take such a risk for someone else? And why would that person risk sharing their culpability with a "rogue engineer" when I am sure they could phrase things in a way that sounds more innocent and reasonable.

> why would such a "rogue engineer" take such a risk for someone else?

It's well documented that people will often defer to authority, even if it's to do something they feel uneasy about [0]

Indeed, in my own experience as a younger engineer I had a manager sitting over my shoulder once dictating what I should but into an estimate, under my name. Indeed it was fairly innocuous stuff, and I'm sure he meant well, but still ...

In a business environment where you have all sorts of contractors, graduates, people floating through and a management culture of "getting things done" there are so many ways that this kind of thing can happen.

[0] https://www.simplypsychology.org/milgram.html

Sure, it won't be phrased that way and so obvious when presented. But I guess that's my point: Have a conscience, pay attention, and don't build anything illegal or unethical. Management may be sanctioning it now, but they won't have your back once the music stops.

As a rougue engineer I would expect at least an increase in my bonus which is proportional to the one of my chef. There is little reason to be a roughe engineer otherwise.

The usual reason is something like "you'll get fired if you don't".

Civil engineers also have professional associations and certification boards backing them up in most jurisdictions whereas we do not. We don’t have anything to appeal to when we go to our boss with ethical concerns besides, well, our sense of ethics. If the order has already made it down the chain, that sense is probably not shared.

FWIW, software engineers do have professional organizations like the IEEE and ACM, whether or not they back members in any significant way. There are also PE licenses although my understanding is that they are being discontinued for software engineering because basically no one was getting them in that field.

The thing that matters is not whether professional organizations exist, but whether their acceptance is mandatory.

The reason why this works in civil engineering is because the boss or the company cannot overrule the engineer. If the engineer declines to sign off some thingy as acceptable because X, Y and Z must be addressed, then that's it - the company can't proceed until the issues have been resolved; the management can't sign off by themselves, and they can't really even fire the engineer and get someone else to sign off for that because a complaint from the fired engineer about that would result in the process getting an independent review, (assuming the issues are real) the process getting stopped anyway, and the unethical engineer who agreed to sign off despite these flaws losing his/her licence - so that even if not signing gets you fired from one place, it's still the better choice than signing and being unable to ever work in that profession again. "Ethics standards in a profession" mean the standards that the profession imposes on its members even if they and their employer want to violate these standards.

Without effective enforcement, however, the whole process would be useless.

Of course, at least many of the mechanical engineers at auto companies most likely do have PEs. It's not just civil engineers. It's mostly any engineer/engineering manager who signs off on drawings, test, etc. for regulators. When I worked on oil rig designs, I started down the path to getting a PE and it would have been expected had I stayed in that business.

That said, I'm sure there would be lots of complaints if software had a similar system given that it would likely come with specific educational and other specific requirements.

Does all that happen often?

I don’t think it happens often.

Though Lawsuits are not uncommon in civil engineering. When I worked as an EIT (passed the state engineering exam, on way to becoming a Profesional engineer), our office did an analysis on someone else’s slope stability calculations, because the slope wasn’t stable and they wanted to figure out if it was done properly.

Civil engineering designs have a clear paper trail and the buck stops with the engineer who signed and stamped the drawings.

In many cases it is "you can over rule once", because that's all you're going to get. So no, it doesn't happen often, but the engineer legally has that power.

The implementation of this cheating is more insidious than you think. There could be a requirement to put in a "simulation" for testing. There's nothing unusual about implementing this feature. Then another request comes in at a later time to make this feature a build setting. Then at a later time someone who is not a software engineer toggles this build setting on production build during manufacturing to put it out in the field.

Hear hear. When building a bridge it's obvious when they're building a test bridge or a real life bridge for people to drive over.

"So you want to spend 100 million dollars on a bridge that you're only using for testing purposes. And you want to put this bridge right where it would be really useful to put a real bridge to handle rush hour traffic. And you have a deal with the city to collect toll payments for this 'test' bridge for the next 50 years. Yeah, I don't think we're going to be cutting any corners for the 'test' bridge."


"You want me to take 2 hours to add in a test feature that can be conditionally compiled in for testing purposes? Also I talked to John in testing and they actually want this feature for testing. Seems legit."

Two months later. "Hey, I found this test feature that would allow us to cheat on the regulations. All we have to do is set this compile flag to true. The programmers even made this neat UI that allows us to change the build configurations without having to talk to them."

In addition, I think it's much easier to go back into the field and examine a bridge to ensure that something important was constructed to the specifications that you required. This is not the case with most of the software people write, certainly in this case it's pretty difficult for a developer to pull a car of the production line and verify that the code they wrote made it into the car without any changes.

Counterpoint: Software engineers can just make it a configuration option and leave it to the user to decide how it is run. Unlike civil engineers who can't build a bridge with a configuration option for a 'light & unsafe' bridge, software engineers can make everything an option, and that's often the best practice.

Of course, assuming that they are unaware of how their user / client / employer will actually use the software, they should be fine.

"It should be noted that no ethically-trained software engineer would ever consent to write a "DestroyBaghdad" procedure. Basic professional ethics would instead require him to write a "DestroyCity" procedure, to which "Baghdad" could be given as a parameter."

--Nathaniel Borenstein

It's kind of hard to disagree...

True, no ethical engineer would write "DestroyBaghdad", but how about perfectly reasonable "set custom threshold for test" function that lets the client pick? And then the client chooses an unethical value.

I can't force my ethical standards onto an unethical employer without the help of a union, or at least a professional organization able to advocate for me.

Employer: Do this vile thing.

Me: No, that would be unethical.

Employer: Oh, you're right, of course. Train Steve instead.

Me: Okay.

[later] Me: Steve is trained.

Employer: Great. You're fired. Steve, do this vile thing.

There is another option.

Employer: Do this vile thing.

Me: No, that would be unethical.

Employer: Oh, you're right, of course. Train Steve instead.

Me: No. <quits>

So you always quit when asked to train someone? The reason Employer plays along in the beginning is so you don't realize Steve will be your replacement until it has already happened.

Besides that, some people will need to continue working as they search for another job, and training Steve could be parlayed into a rudimentary supervisory position on the resume. That helps if you ever intend to jump over to the developer-manager track.

Q: "Have you ever had anyone under you?"

A: "Yes. I was in charge of a trainee once."

Q: "How did that work out?"

A: "Great, until they gave him my old job, then fired me. Maybe we had a budget problem?"

Additionally, if you don't train Steve and instead just quit, that's not going to stop the Employer from asking Steve to do the unethical thing without training. Now the unethical thing is not only unethical, but it is also filled to the brim with bugs that cause unsafe situations for end users.

For what it's worth...


"The program is formatted to recognize employees who contribute ideas, or those who raise questions about safety issues before they become bigger problems. Call it an internal whistleblower program."

What does this mean in the VW case? I believe the unions are strong in Germany. So either the "rogue engineers" was a lie from the management or they were doing it in their own volition?

don't give up so easy. Think of Fair Trade coffee -- all it takes is a vocal minority to insist on a set of practices and that will come to be considered the norm. The social (and market) pressure of hiring only coders that have taken the "Oath of Shannon" or whatever will enforce the standards far better than any union.

Fair Trade coffee is not an inherently beautiful and good system. It has worthy goals and maybe does more good than bad, but it's hardly a great example.

It’s a metaphor. Nobody was forced to adopt it with any kind of governmental or economic coercion, just a vocal minority that thought it was a good idea to elevate the standards and do more than just buy the cheapest product in the market.

Wouldn’t fly in most European countries with reasonable labor laws, union or not. The company would be in deep shit.

I would say the U.S. has become blase about corporations not having a social benefit component. They can do really vile things, far worse than a human, and get off with a fine that may or may not be meaningful in inhibiting such behavior in the future. I can't think of a single historical case, which I think would be notable, of a state rescinding the articles of incorporation of a company thereby destroying its shareholders - the corporate equivalent of capital punishment.

Some states do have "social benefit" corporations as a thing, where their charter expressly states how they are to benefit society not merely make profit. But no state makes this kind of incorporation mandatory.

VW is a German company. Nissan is Japanese. VW engineers did their unethical thing in Germany. In the context of this story, your comment fails logically.

Well, AFAIK it weren't the VW people who build the cheating system and the ones who built it, pointed out that in some countries it might be illegal to use it. The VW people just decided (probably a manager) to activate the cheat system.

So I am wondering where an ethical standard could help in that chain. If you build international products its kinda hard to know every law on this planet and refusing to build something, just because someone could use it in a country where it is illegal to use it, could become very complicated very quickly.

I think the simplest solution would be to improve the test methodologies and maybe increase the penalties.

The cheat device that the VAG group added wasn't really in any sort of grey area.

This isn't like the wrong shape steering wheel, or accidentally too small wing mirrors.

This was a deliberately programmed system for cheating on emissions tests.

These things are pushed by management down, rarely from engineers upwards. The usual "solution" is to hire people from different cultures with lax laws, i.e. in WEU it's popular to hire EEU folks, in USA to hire Indians to do the dirty job, as they either don't understand the legal consequences or are from cultures used to "work around" laws.

> but active in fraud.

No we're not. We don't have societies that put any such responsibility and restrictions in place, along with benefits, for what we do. Whatever the company does with the code is the sole responsibility of it.

Like you said, other engineers have LEGAL obligations, what makes an obligation moral or not is extremely debatable and always subjective to the point of uselessness.

I'm not entirely unconvinced this isn't already illegal somehow. There's a lot of laws. I bet if you dig through the definition of "certification" for emissions standards you'd find some reason why everyone knowingly involved with a fraudulent submission is already subject to some legal penalties somehow already.

"It's hard having moral obligations and shit so why even bother with that"

Are you suggesting that ethics and morality are useless, and that software developers should not care about them?

No, he's saying that if he raises an objection, then his boss will fire him and find someone less competent to do the job. Then after he's fired for standing up for the good of society, society will tell him to starve to death because he doesn't have a job. Finally, his replacement will do a significantly worse job and the result will be greater damage to society.

This is why unions exist.

IANAL, but from what I can tell, unions tend to be organized primarily to protect workers, not the general public. Refusing to do unethical work is usually legally protected (e.g., by whistleblower laws), so your union representation can help you out (they have lawyers, and experience dealing with labor law) but it's not necessarily their main purpose.

When has a union in this context ever protected a member who refused to take action on behalf of the business for ethical reasons? Lots of unions protect inaction motivated by worker safety, but this is a different thing.

The -idea- of unions. They're not exactly in a strong place in the US.

I'm not. I'm saying that even if I were to be protected for standing up for immoral practices (which by and large I'm not) we're still putting the burden of morality in the wrong entity, the employee at the bottom of the decision tree. This is a burden that the state needs to impose on the top of the company directly, not have it be shifted to the bottom.

My comment, the OP, is not saying that the bottom of the tree should have the legal obligation. But I would argue that every branch should consider their own ethical responsibility. Don't confuse those two responsibilities.

Ethics isn't that simple. The tools need to exist and some people will use them for things we don't agree with.

The emissions software needs to be able to turn things on and off because the emissions requirements for certain jurisdictions and applications can be mutually exclusive. It just so happens that manufacturers flip switches they souldn't in order to fake compliance.

I've worked on software that can be used to make it much easier to drop a bomb on you. The main purpose of the contract was a software package for defensive applications (i.e. making it much harder for a car bomb to surprise a checkpoint). Should it have not been built because it has the potential to do things you or I might not agree with?

This is easy to pontificate about. How do you actually bring about culture change?

This is a great question. I would agree a lot with gnulinux here. I have a degree in physics and took some engineering class. In both the engineering classes and physics classes we talked about ethics, moreso in the engineering ones. We also had a required ethics course.

But I will add that another thing to help is to encourage it within your social sphere. This can easily be done by just asking your colleagues about their ethical concerns. You can also encourage this by talking about things that aren't your software. This helps everyone get in the frame of "I need to constantly think about ethical implications". Other people are talking about fear of being fired. While I think that has to come down to you and your ethical code/leverage, it is much harder to fire and replace a team. So I would actively encourage this kind of talk. Having worked as an engineer I can tell you that we constantly did talk about ethical implications of things that we worked on, things that our technology could be used for, and if other companies were doing ethical things. It's hard to draw the line sometimes, but the best thing I can think of is to constantly be talking about it and being aware that you have power.

Education. Engineering and science schools should have good and rigorous ethics courses. Taught by actual philosophers and engineers on the field dealing with ethically sensitive issues (like writing a real-time OS for an airplane, self-driving car etc). An engineer or a scientist shouldn't graduate without the ability to reason about the social and ethical implications of their jobs. Engineers, physicists, scientists, social workers change the world more than most individuals and thus they need the ability to reason ethically more than most individuals.

I think the real answer here is "legal obligation". You may sway some number of sw devs with classes on ethics, but in reality many will throw it all out the window to save their job unless there's a real and significant consequence in breaking said ethics.

You're mostly right. As I said to some other answer below, I think law enforcement is crucial to limit evil people doing evil. But well-intentioned people can do evil too, just because they didn't know any better. Engineers who made Therac-25 probably didn't wanna kill bunch of people, but they did. (Also read Hannah Arendt on how she talks about most Nazi officers thought 'they were just doing their jobs') Understanding that your job may have such implications is why ethics classes are very important. This shouldn't be optional, engineers should have the ability to be able to think about the implications of being an engineer.

Of course, I wonder if that would result in no-degree-me being out of a job :). I have worked in medical equipment and cancer diagnostics for 14 years now.

Well, I didn't mean to be an elitist. My point was that ethics is an important "skill" for engineers and you can train this skill by other means such as reading books. There are tons of books about theoretical ethics, scientific ethics, engineering ethics, bioethics etc. Even reading something as basic as Stallman can be very helpful for software engineers. I'm just pointing out that engineers should think more about ethics.

>Well, I didn't mean to be an elitist

No no no, I didn't take it that way. I was just remarking that I may find myself out of a job if regulation similar to civil engineering existed in our field due to the fact that I don't hold a degree. Nothing beyond that and it wasn't a response to anything you said.

I've argued to (and with) my coworkers on more than one occasion that we _should_ be personally liable in certain situations (we already have the FDA watching over us at a high level.) For example, if I were complicit in fudging the results of an inter-system repeatability test in my current position.

Do you think that teaching an unethical person about ethics makes them ethical? I never understood the point of ethics classes.

It’s hard to imaging that people working on emissions cheating would have been less likely to work on this, had they only taken the CS ethics class.

It's not about making an evil person not-evil. You don't have to be evil to make evil things. The point of ethics classes is to learn how to think ethically. How did old and modern philosophers think about ethics? If you're an evil person, then law enforcement should be there to discourage you not to be evil and hopefully penalties are high enough to stop you. But if you're not an evil person, and the law gives you certain freedom on ethically sensitive issues, then an ethics class will help you how to reason in such situations. When I don't debug my code properly sometimes I feel terrible because a bug can significantly change a person's day or maybe even injure them (I work in telematic industry). Being a competent engineer requires one to understand the implications of one's job and act accordingly which may be debugging properly etc...

Certainly not. But they can help sway people who are on the border to the ethical side of things. Ethics classes can also provide tools and mental models for dealing with gray areas which aren't as cut and dry as cheating on an emissions test.

Yes, and the the thing is that most people are "on the border" - that is to say, little more than a product of their environment (speaking of their actions/attitudes, not their worth as people). If you teach an entire population ethics (assuming it's done well and everything that comes with that), you will still have your 5-10% of rotten apples, just as you'd have 5-10% who would seek out ethics on their own. The big difference will be in the 80-90%.

Like most classes, the point is to educate people about something they don't know about, broaden their horizons and perhaps cast new light on apparently familiar topics.

I've never understood the point of computer science, I can produce a perfectly functional BASIC program.

Ethics has its own vocabulary, history, and arguments in ethics are common and often don't have a resolution other than compromise. It's predicated on making logical arguments and has ties to epistemology: how do we know what we know?

So saying you don't understand the point of ethics classes, as if people are ethical or are unethical, is like saying you don't understand the point of teaching logic or debate as if people either pop out of childhood with an innate conformance or resistance to one kind of logic or debate, no other kinds exist and there are no arguments.

If you took an ethics class and don't understand the point of it, then I'd argue it wasn't a good ethics class.

> We are writing things that affect peoples lives. That could kill people. That affect the environment. We need to hold ourselves to the same moral obligations as other engineers. We have a duty to prevent harming people when we write our software.

High and mighty ideals, but come on.

If you're working on software for a medical device, then yes.

But on another post today it was pointed out that crypto currencies are burning more electricity than Iceland.

That's straight out environmental vandalism. Is everyone involved in crypto expected to down tools, cos ethics?

Nobody told me about the code of ethics I have. Can you send me a link? Is it somewhere in the windows 7 EULA I can never get myself to read those darn things :)

Well in this case it was factory testing and had nothing to do with software.

In the case of VW, I highly doubt those software engineers were told plainly what they were doing. While I agree they have an obligation also keep in mind speaking up would probably be career suicide.

The liability is on the executives that make take these illegal actions.

I understand your sentiment, but I think this only applies if you want to be an especially good person. The rest of the professional world does not have to match such a high standard of ethics, and as much as I would love for everyone to have to behave nobly, it's pretty unreasonable to ask software engineers to report software that exploits loopholes in regulations while our counterparts in banking, private equity, law and medicine basically live exclusively off of doing that.

edit for the downvoters: in this exact case, the executives of the companies hacked the system, their lawyers probably approved of it, and the strategists who conceived of it likely used to work as investment bankers. And yet, the software engineers are the ones expected to take the high road?

> while our counterparts in banking, private equity, law and medicine basically live exclusively off of doing that.

Literally every one of those professions are governed by legal ethical requirements. Software engineers/devs are not, which is kind of the whole point.

PE doesn't any more than programming. Banking's laws are just moats for the established players masquerading as regulations.

Law has them, but in a strange twist, their ability as lawyers make it so that they don't matter. And medicine is too complex for me to really comment, but from my vantage point, it looks like they're free to be scumbags so long as they claim they won't be when they get their degree.

But every one of these are largely just regulated by the idea that if you're too big of a scumbag, the feds will nail you for wire fraud. That seems to be the only real rule of law or code of ethics we have in the professional world. And software engineers have to obey it as much as everyone else.

>>PE doesn't any more than programming. Banking's laws are just moats for the established players masquerading as regulations.

PE sure, I probably should have noted I know little about that, but when you talk about banking you're now going from "have no ethics laws" to "well, they exist, but they're not effective in my opinion."

>Law has them, but in a strange twist, their ability as lawyers make it so that they don't matter

Lawyers get busted on ethics violations all of the time. My mother is a lawyer and for a short time I worked as a pseudo legal secretary in a mid sized city. I recall numerous lawyers being disbarred for ethics issues. I have no idea what the real numbers are, but I don't think you do either.

>And medicine is too complex for me to really comment, but from my vantage point, it looks like they're free to be scumbags so long as they claim they won't be when they get their degree.

So you don't really understand the issue, but you've heard some stuff on the news and maybe from friends, so you write the entire things off.

How about civil engineers who are personally liable for their work?

>But every one of these are largely just regulated by the idea that if you're too big of a scumbag, the feds will nail you for wire fraud. That seems to be the only real rule of law or code of ethics we have in the professional world. And software engineers have to obey it as much as everyone else.

Again, moving goalposts. If you don't feel the current regulations wrt doctors, lawyers, and bankers are enough, fine, many would agree, but that's not the issue here.

>PE sure, I probably should have noted I know little about that, but when you talk about banking you're now going from "have no ethics laws" to "well, they exist, but they're not effective in my opinion."

Oh, if we're going to be pedantic about what exists, then software engineers have them. There are computer hacking laws! And software IP and copyright rules. See, we're regulated, too. I thought we were operating on the premise that the existing structure was insufficient or irrelevant to any meaningful end, but since it technically exists, I'm not sure why you're arguing that we need to create them.

Honestly, your entire reply seems like you're going out of your way to miss the point. When you made a mistake about PE, I didn't concoct some strawman to try to embarrass and highlight your ignorance, so I don't know why you're trying to do that to me. A software engineer's counterpart isn't some ambulance chaser getting disbarred. Biglaw associates are the professional counterparts to software engineers. And in medicine, I've seen tens of thousands of extra opioid deaths and hundreds of billions in insurance/medicare scams attributed to the medical industry to have a bit of an opinion at least.

If you think the argument that industry-specific regulations don't matter, and that the only regulations that matter are federal fraud laws is a goalpost change than I don't know what to tell you. My argument is that the industry codes don't matter, and that any semblance of actual regulation takes place at the federal fraud level, so putting an unnecessary step of internal regulation is pointless.

The general assumption at the time when VW was put under a microscope for these issues is that everyone is cheating, VAG just got caught first.

What stinks seems to be the way the regulations are being implemented moreso than what sort of detours the manufacturers can take to avoid them.

I won't be surprised when we see more auto makers added to the list, however I wonder if anyone will criticize the regulations instead of shaming auto manufacturers.

Here's a non paywalled BBC article: https://www.bbc.com/news/business-44763905

> VAG just got caught first.

Sure. Since the VW incident, many cars from several of the big automakers have been found to fail emissions tests.

> I wonder if anyone will criticize the regulations instead of shaming auto manufacturers.

It's telling that the automakers have chosen to cheat rather than call out the tests for being unrealistic -- or build a slightly less powerful car.

By doing so, they forfeit any possible sympathy from me. You don't get to cheat on a test, brag about how you passed, and then if you're discovered, claim that the test was too hard. Too late.

Seriously- the Jetta TDI made 11% less power[1] in compliance mode. This isn't about the regulations being hard to meet, it's about flagrantly disregarding the EPA. It's about how pathetically toothless the EPA is, that they didn't have the funding or reach to discover this and instead college students figured it out.

[1]: https://jalopnik.com/heres-how-much-horsepower-volkswagens-l...

The year I bought my Jetta it was the most efficient midsize sedan by a wide margin. I bought it because my friends had Priuses and it was feeling pretty monochromatic at my house.

The skeptic in me should have known better. One manufacturer was blowing everyone else out of the water? German engineering? A seriously nifty gearbox? Are those enough to explain the difference? Apparently not.

Toothless? The EPA is like a great white shark compared to European regulators.


Yeah, the fact that Scott Pruitt was the head of the EPA for over a year was a sick joke.

It's not even like the test couldn't be met, VAG just wanted to skip the urea system. At the same time as they were cheating, they were selling other diesels with urea that met the regs just fine, no cheating.

The test could even be met without the urea system. But the car would have to run rich and flush out its NOx system more often, which uses fuel and lowers performance. The thing that was so marvelous about the Volkswagen TDI engine was how all of its performance numbers and fuel economy were better than what was advertised. Other manufacturers were racking their brains trying to figure out how they did it.

It seems a little hard to believe that other manufacturers had no idea how they did it. If it was so confounding it could quickly resolved by plugging in an analyzer and run a test cycle...

Maybe VW isn't the origin of this, they could have seen another manufacturer doing something similar, decided to join in and took it too far.

They may have done exactly that, but not in highway conditions, and so got fooled just like the inspection locations.

What's the point? Urea is almost free. I don't get why it was ever a big deal.

The DEF itself is pretty inexpensive (though not free, like $8/gal), but the system to keep it from freezing in wintertime is actually kind of a pain, since it freezes at -10C but decomposes around 60C so your heating element can't be too hot. The cost savings from not using it were rather substantial. Volkswagen TDIs were the cheapest performance diesel cars by a rather large margin.

It's an active expense that doesn't directly benefit the user, and disables your otherwise mechanically sound vehicle when it runs out; both of which tend to rub people the wrong way. Also I think there was something about german companies colluding over adblue tank size to force more refills, but don't quote me on that.

Also it's a user-maintenance thing. One more thing to refill on top of your windshield wiper fluid.

I forget if it was VW or Mercedes, but one of them branded theirs as "blue-tec" or something that doesn't sound like you're handling piss bottles. Let the mechanic worry about that.

In all fairness, the urea systems are really annoying. I was driving a rental vehicle and I got a warning that it was out of urea in the middle of nowhere. I stopped at a gas station which did not have any and the vehicle would not start until you put urea into it which I could not find in the middle of the desert.

I had a co-worker who had a similar horror story. Drove out into the Alaska outback in a modern truck (<3 years old), there was some issue with the emissions reduction system (DEF heater failed), the vehicle went into "limp" mode after 50 miles (40 MpH speed limited) and displayed a warning it would limit them to 20 MpH after another 50 miles.

There literally wasn't any dealer or provider that could do anything before it hit 20 MpH limit, and the truck wound up being flatbed-ed (luckily the manufacturer picked up the bill, but still).

This definitely makes diesel vehicles less reliable, more expensive to maintain, and likely reduces their lifespan. If I purchased a truck I'd go gas.

I imagine that this is when I would perform a hard reset (by disconnecting the battery or pulling the fuse) of the ECU.

Not sure if it would work on that vehicle, but it generally clears emissions codes and the like, even on fairly recent vehicles. Its possible it might have let the car drive another 50 miles or so.

I don't have any evidence on hand, but I distinctly remember most of the auto makers complaining every step of the way about most emissions regulations.

Granted, it tends to get a "boy who cried wolf" feel to it as they've always done it, but they absolutely raised a stink about the constantly rising regulations.

  “It’s sort of a case where the dog actually caught the bus,”
  Gillis told me. “Now what’s it going to do?”

Referring to American car manufacturers' anti-CAFE wishlist being adopted by Scott Pruitt's EPA. The wishlist was submitted as a sort of bargaining piece, the idea being that industry demand something extreme expecting the incoming administration to meet them halfway. Except Pruitt just accepted it without reservation[1], and the industry was caught flat-footed. So now they're actually _repudiating_ some of their arguments.

The dirty little secret of the car industry is that efficiency and safety regulations help drive the upgrade cycle. When every car from this year has feature X, then all used cars instantly become out-dated and even unsafe in the minds of consumers. And the timelines set out by the regulators help to coordinate competition in a way that would be illegal otherwise. For example, if a new efficiency or safety standard is still 5 years out, then everybody knows to a high degree of certainty how each competitor will time and stage their rollout (not too early, not too late). The benefits may not exceed the overall costs of the regulation, but the industry is built around the system and pulling the rug out creates chaos.

[1] I guess this was a case of the Trump administration taking industry's rhetoric literally and not seriously.

Like you said, everyone is going to cry when their burden is increased in any way, even if it's a completely reasonable increase. Companies complain if their employees request cost of living raises, and people complain when prices rise due to rising cost of living. At some point, it drowns out any legitimate complaints. If you've been saying the regulations are too high yet have been meeting it every time, then a new regulation comes out that is actually too high and you can't meet it, people will take you less seriously.

If you then pretend to meet those requirements even though they were too high, you get zero points.

"It's telling that the automakers have chosen to cheat rather than call out the tests for being unrealistic"

They do that, too. You can probably imagine the response the HN crowd would give, which is pretty much the same response the news media would spin it with... money-grubbing rich companies trying to escape from beneficent, unambiguously good laws so they can make more dirty, polluting money. It doesn't play well, so they don't make this argument in public. But if you live in Michigan, you can find plenty of engineers who grouse about the standards in various ways, and not just "they're hard". (Believe it or not, engineers have families too, and like them and their families to not die in the cars they quite likely drive themselves.)

Between the media's drive to drum up antipathy towards the automotive companies for both their own ideological reasons and simply because stories about big corrupt companies sell, and the ever-present and unavoidable politics (of all kinds) involved in any sort of standards setting, there's so much noise that on those occasions where literally impossible regulations are created, the system is not very good at dealing with it. I am speaking neutrally here, because in the real world, balancing between costs and benefits of regulations is a very hard problem and I'm not really yelling at any particular party here for not getting it perfect. It's a systemic problem.

I rather strongly suspect that if we could get a perfect judge that the current set of regulations is probably already impossible to comply with at all; failing that it probably can't be complied with economically. I seriously doubt emissions are the only problems, they're just the ones people are motivated to test. I bet if Consumer's Reports or a similar organization set up a full legal compliance suite they couldn't buy a single car from a dealership lot that complied completely with everything. I think they'd be close on most measurements. But I bet every vehicle would fail something, and there'd be a non-trivial number of rather substantial failures.

(For an example of the latter, noise regulations on how loud cars can be as they pass you at various speeds are really easy to fail on a specific car. A particular bit of trim that came loose during shipping or something can fail you out very easily. Now, this one is less important in the grand scheme of things than emissions, but it's an example of a regulation that could pass at the test site but fail on many or even most cars in the field.)

Considering that, in recent years, cars got larger, more powerful, have better mileage, and pollute less: I think the regulations are extremely realistic.

It's not like that late 1970s and early 1980s when cars became smaller and had anemic engines in order to meet with new regulations.

"Considering that, in recent years, cars got larger, more powerful, have better mileage, and pollute less: I think the regulations are extremely realistic."

You are assuming away the very question I'm asking, though; are those cars actually in compliance? If not, you don't have a point. (I don't mean "you are wrong"; I mean, you literally wouldn't have a point to your post.)

I didn't say anything about cost, reliability, complexity, ect.

I don't even know how to square that reply with what I said, to the point I wonder if you meant to reply to another message?

> You don't get to cheat on a test, brag about how you passed, and then if you're discovered, claim that the test was too hard. Too late.

It is the competition that pushes you to such extremes. If others are doing it, then what is to stop you.

Too much regulation and court procedures are preventing others from coming out and blowing the whistle on themselves.

The point of regulation is so that everyone has to reach the same bar. Breaking the rules to cut costs doesn’t absolve you.

And when the regulation fails to regulate for a long period of time, such that all eventually meet the regulation with either a loophole or a crime, it does not make the system any fairer when the axe falls.

If the reg's were never put into effect, and companies never given citations, it is institutional failure. And very unfair to all entities who are suddenly scolded at not meeting the criteria.

VW presumably started it. Other companies eventually competed. And all of a sudden everyone is now in violation.

It is like cheating on a test. One kid starts it. Teacher does not care. Kid does not get scolded. Competition gets intense. Rest of kids start cheating too. The axe falls, sometimes on the dull, slow kid, and then everyone falls in line. But then teacher goes back and sees who else cheated in the past, and the entire class is outed. How is this fair again?

> Too much regulation and court procedures are preventing others from coming out and blowing the whistle on themselves.

"The penalties for murder are too harsh, so people aren't confessing to many unsolved murders."

Well, yeah. I would not blame "the system" for this, however.

Murder and emissions fraud are like apples to oranges. Emissions fraud are a corporate practice, while murder is a one-off event. Coming clear on the latter improves all industry.

For the record, there have been cases where criminals have come forward for crimes after statute of limitations expired. I am not sure it if was in particular to murder though.

I think you get my point, though. Amnesty for rule violations does not discourage future violations. "Everyone's cheating on emissions testing and there's no penalty" isn't going to dissuade anyone from cheating on emissions testing.

You can compare it to "you are allowed to murder anyone you want, as long as you confess afterwards" and see that that's not the best public policy if your goal is to reduce murders.

> however I wonder if anyone will criticize the regulations instead of shaming auto manufacturers.

What's wrong with the regulations? It seems that they're designed to prevent cars from spewing toxic shit into the atmosphere, something that seems to be an admirable goal.

This mindset is exactly the problem: you’ve judged the regulations on their goal, rather than their effectiveness and how they interact with the industry’s incentives.

Well what do you suggest then? I mean, the current regulations seem to have caught a number of car manufacturers faking test data and presumably they will be punished.

I think we can agree that the goal is to reduce dangerous emissions, but if we're not going to have regulations that state a limit and perhaps a penalty for abuse, then how?

The worry is that they start to become impossible (or at least impossible without massively increasing the cost and complexity of the system).

So "cheating" starts to look more and more worth the risk.

Granted, I don't think we are quite at that point yet, and don't take this to be implying that we should not have environmental regulations here, just that we need real information about what's possible and how, and there needs to be a 2 way conversation with the manufacturers to evolve the regs with what's possible.

> The worry is that they start to become impossible (or at least impossible without massively increasing the cost and complexity of the system).

That's not really a worry. Governments are well aware that people need cars that they can afford. Governments are ultimately elected by the people, and messing around with transport makes you unpopular (the recent backlash at the possibility that fuel duty will be unfrozen in the UK is an example).

Governments don't care about people having affordable cars. Governments care about votes. Look at the cost of safety features that have been implemented in cars the last 10 years; some for marginal benefit (rear-view cameras). These features increase the cost, making it hard for lower income people to afford the cars.

> Governments don't care about people having affordable cars. Governments care about votes.

These things are related. If the people think that they can't get the car they need because of the government this will turn in to loss of votes. See fuel duty example again.

> Look at the cost of safety features that have been implemented in cars the last 10 years; some for marginal benefit (rear-view cameras).

Regulations are rarely perfect. It seems from my searched that backup cameras are net better (~300 people killed, 18,000 injured in backup incidents) and that cameras will help drop some of those figures, but not by a huge amount. I guess we'll just have to wait and see how the figures play out in the future.

However, some regulations have had a significant effect, for example seatbelts.

> These features increase the cost, making it hard for lower income people to afford the cars.

The tech will come down in price, and the car companies still want to sell cars to lower income families and so will at worst absorb some or all of the cost at the price of a slightly lower profit.

And as someone else pointed out in a different thread, emissions from cars has a significant negative effect on people's health, particularly in urban and inner city areas where these lower income people are more likely to live.

Mr Fox, I need help understanding how to securely design this henhouse.

>I think we can agree that the goal is to reduce dangerous emissions,

we can reduce the emissions to zero by banning all cars, would you support that regulation?

>then how?

Improvements to technology the lessen the dependency on those energy sources.

Regulations will not solve the problem, no matter how draconian you desire to make them. Regulations will just increase the cost of goods furthering the economic divide and harming the people that can least afford the changes

> we can reduce the emissions to zero by banning all cars, would you support that regulation?

That's ridiculous, of course not.

> Improvements to technology the lessen the dependency on those energy sources. > > Regulations will not solve the problem, no matter how draconian you desire to make them. Regulations will just increase the cost of goods…

Regulations will drive technological solutions, and if it weren't for regulations around emissions then car manufacturers have proved that they will quite happily sacrifice the health of everyone in the search of profit.

> furthering the economic divide and harming the people that can least afford the changes

Oh please. Car manufacturers will always sell budget cars to people with small budgets, because there's a market for it.


Please don't post ideological flamebait to HN. We ban accounts that do that. We've also had to warn you about this before.


I don't actually think you're interested in that list, but there is a whole theory with scholarly support describing how regulations inspire innovation. It's called the "Porter Hypothesis" (the Porter of Porter's Five Forces), and like all theories it has some evidence for and some evidence against.


Very interesting -- I had no idea this was a scholarly domain of study; thanks for posting this.

Here's a survey that might be interesting to you: https://www.itif.org/files/2011-impact-regulation-innovation...

I can't promise that there isn't some bias in one direction or another, I just found it interesting, and the conclusion is very pithy.

> I think your definition of "budget cars" and mine are vastly different, Regulations have already priced new car ownership out of the hands of many people.

So buy second hand perhaps?

I bought my car for £5000 second hand (yes, I can afford a new car, but why bother, they depreciate so fast in the first few years it hardly seems worth it) and it's a nice car. It's way more than I could have paid for a perfectly serviceable second hand car. I checked, £700 was the cheapest of the same manufacturer as mine, and it's 5 years newer too.

>>> quite happily sacrifice the health of everyone in the search of profit.

> All yes, the evil for profit corporations...

Wait, so it's OK to pump shit into the atmosphere as long as you make a profit?

Is the bar really so low for communism that believing that we shouldn't be poisoning large amounts of the population gains you membership?

Corporations and companies are optimised for one thing, profit. They do this at the expense of the climate and the health of people. Hell, this is a comment on an article about how a car company cheated emissions tests so they could sell more cars. Emissions that kill people.

Do you really believe that we would have the current emissions levels if there was no regulation?

> Well comrade

Oh please drop this shit. Resorting to name calling just suggests that you don't know what you're talking about or have no real point.

> show me all the innovations that regulations directly have inspired. I would be interested in the that list.

If you lower emissions limits over time car manufacturers will have to innovate to meet those targets, right? I mean, car companies aren't going to do this themselves are they, see cheating on the emissions tests, again.

> I always forget how must support for communism there is here in HN circles...

I'm not a communist.

The other side of the coin:

In constant dollars, the average cost of a new car from the beginning of the emissions & safety-regulation era (say, 1970..) has about doubled. This is a nontrivial chunk of income. As a brit, you might live someplace where cars are optional. Many Americans have no such choice...

Likewise, the low-hanging fruit has already mostly been picked. A new 1964.5 Ford Mustang, with the engine turned off, sitting in a garage, releases more toxins than a modern car does while traveling at highway speeds. The 90/10 rule absolutely applies, as it does to safety. We're coming close to having to install catalytic converters inside the engine itself to minimize the lightoff time, which is about the only place you can find NOx/VOC emissions to clean up nowadays.

Not only were there upfront costs to recover for emissions & fuel economy, there were the knock-on effects on performance and reliability that took the better part of two decades to solve.

> I always forget how must support for communism there is here in HN circles...

I always forget how libertarians think that anyone who holds views that don't completely conform to their ideology is necessarily a communist (or "collectivist").

Have regulations not solved water potability, domestic child labor, lead in paint, water tables, airline safety, accessibility, and many more problems?

You will disagree, but largely no. Those things where solved by a variety of other factors and where on their way out long before the government came along to "regulate" them and take the credit for it.

Could you elaborate in some detail on those other factors? Preferably with some citations. I’m totally unfamiliar with this second narrative, while I am familiar with the likes of ‘The Jungle’.

> Regulations will just increase the cost of goods furthering the economic divide and harming the people that can least afford the changes

NOx gas emissions from cars cause and worsen respiratory diseases, especially in urban areas where it is especially concentrated, thus ALSO harming people that can least afford it.

I'd also note there are plenty of cheap vehicles that do meet these regulations.

> you’ve judged the regulations on their goal, rather than their effectiveness and how they interact with the industry’s incentives.

The regulations were ineffective because the EPA is unable to do its job. They test cars on a dyno in a climate controlled room instead of in the real world. They are embarrassingly vulnerable to defeat devices because they don't actually try to find defeat devices. Shitty nominees, corporate lobbying, and frivolous lawsuits have neutered the EPA from the inside and outside.

Care to explain how they're ineffective instead of repeating libertarian platitudes? Is your argument that they are ineffective because manufacturers are cheating on the tests? Seems like the ineffective part here is not the regulation but the enforcement. These regulations are expected to be misaligned with industry incentives. That's why regulation is necessitated.

Please don't do ideological name-calling here, regardless of how wrong another comment is, or feels.


There are no more coral snake antivenom manufacturers anymore because the regulations became more stringent and the company that made the only antivenom that was grandfathered in under the old rules decided it wasn't worth it anymore.

Realistic regulations are good because they prevent resource rich companies from externalizing their failures onto others.

Unrealistic regulations cause industries to give up on things that our population needs to survive.

For example, in the past if you got bit by a coral snake you could get antivenom. In the present, over ambitious regulation says that you die when you get bit by a coral snake. Because the alternative to death (using under regulated antivenom) is apparently worse?

If you want to you could regulate that cars run off of happiness and that their exhaust contains no pollutants. But the result would be that either everyone cheats on the tests OR that there would be no more cars. Including the trucks that bring us food and other useful products.

I mean, we're not talking about all regulations here, we're talking about the emissions regulations. Emissions that kill people.

So I don't really see the point of the coral snake thing, interesting though it is.

> If you want to you could regulate that cars run off of happiness and that their exhaust contains no pollutants. But the result would be that either everyone cheats on the tests OR that there would be no more cars. Including the trucks that bring us food and other useful products.

Well, I'm not suggesting that at all so again, not really relevant.

I'm suggesting that the emissions regulations that we have are fine. Cars can and do pass these without cheating so they're clearly not unrealistic.

I think the implication is that the regulations were poorly implemented and didn't test true emissions. I don't see how anyone can view these stories as an indication that emission regulations are a bad idea or somehow aren't needed.

Some people have tried to argue that the cheating is definitive proof the regulations were too hard, and so obviously emissions regulations should all be scrapped.

I really don't understand how one can make this argument. No one argues that we should do away with taxes completely just because some people cheat and pay less than they should.

If anything, widespread cheating shows the importance of regulations. It shows that companies are willing to lie to both regulators and the public. For something that is extremely difficult for the consumer to test themselves like emissions, it shows that an independent regulator is required to keep companies honest.

As far as I can tell, it's pretty simple- they've always been ideologically opposed to emissions regulations, so every headline is bent to accommodate that perspective.

VAG deliberately programmed around the tests (a defeat device). It's not that VAG sent a car to the test but the test just didn't test regular driving VAG deliberately changed the results for cars under test.

I meant "true" as in "what actually happens in the real world" and not "the opposite of fake". You are certainly right that the emissions themselves weren't faked, just that the car behaved differently in the tests compared to what happens in the real world.

In the Seattle area, cars made after 2009 aren't required to have independent emissions tests - the car's ECU is trusted to give an accurate result.

To test cars made before 2009, they put a probe in the tailpipe. It's like the regulators have sold out, and gone full capture.


Do they even look at the ECU on a car after 2009? California trusts the ECU on most cars after 2001, but they do plug into it. Seems kind of BS to me, at least there should be some randomized real testing, and if those don't look good, some models get blacklisted from ODB-II testing.

The state emissions tests are only meant to catch cars polluting above something like circa-1980 emissions limits. Even flagrantly cheating, a modern car is no where near dirty enough to fail the tailpipe check.

The regulations don't take into account market reality. People don't want to buy cars because they're low-emission, they want to buy cars because they're fast, big, and safe. They need to be bigger, faster, and safer in order to compete. But lowering emissions makes that much more difficult. So of course you're going to lie on emissions. If you don't, your competitor will, and then you lose market share.

What would regulations look like that took market reality into account?

The whole point of regulations is to go against what the market will buy without them. If people bought low emission cars without any outside forces then you wouldn’t need any regulations. Of course, in reality that path leads you to smog-choked cities and lots of dead people.

The "Big Cars" part of your comment is really US-centric. In Europe, Japan, and other places, many people want small cars. Some people want faster but most people don't care because there are speed limitations on the roads anyway.

Low-emission might not be a priority, but people still think of it, and some governments measures target the most-polluting vehicles to discourage their use.

Which is why we need make the punishment fit the crime. The problem is, an appropriate and fitting punishment is going to be painful for a large number of people completely associated with the crime.

If we want companies to follow the law, then those that violate emissions standards should face harsh punishments. For example, they could be banned from selling vehicles in the US market for one year, all of their lineup should need recertification under more strict oversight, and they be required to fix or buyback the offending vehicles. And personal liability for senior executives.

The major problem the US has with all automotive regulations is GM and Ford get their lobbyists to write the laws with loopholes that cater to them. For example, CAFE standards that are based on vehicles footprint and are different for passenger cars vs. "light trucks." While the regulations are "fair" on paper, but in practice, they are much more lax for the types of vehicles domestic automakers sell.

Can't you just buy a competitor's car and test it yourself and report them if you find cheating?

And now you just motivated a CarTel. :)

Not if you think you can cheat harder than them.

> People don't want to buy cars because they're low-emission, they want to buy cars because they're fast, big, and safe

You forgot the two biggest factors - cost and efficiency. Most people doesn't want big or necessarily fast cars. Most people want cheap cars that go a long way on a little fuel. However, it's difficult to make cars that are cheap AND fuel efficient AND environmentally friendly.

And now the car manufacturers have got caught cheating, so they're going to have to (I guess) face some penalty, and stop selling cars that cheat the tests.

People might want big, fast and safe cars, but they don't get to have them at the expense of the health and wellbeing of everyone else.

Your competitor is only going to lie if he can get away with it. This only shows a lack of enforcement on the state's part.

Bigger and faster is precisely the trend that should be reversed. While safely systems overall has improved, making faster cars makes them less safe. Cars should NOT be built like race cars. We can have separate race tracks for that.

making faster cars makes them less safe

No it doesn't. Driving a car faster is less safe, all other factors being equal, but simply making a faster car doesn't make it less safe. Look at Mercedes's performance cars, which typically lead the market in safety features. An example off the top of my head is the carbon cone built into the front of, I think, 2000-era SL500s. They could dissipate a tremendous amount of energy, making that fast car significantly safer than most other cars on the road at the time.

Aside from exotic safety features, performance cars tend to be better weight-balanced and tend to have better brakes and better tires than average cars. In normal driving, performance cars are safer because they'll be easier to control in emergency maneuvers.

(Of course when you push the limits of the car, all bets are off again.)

This isn't true for the SL500 at all, and I would venture to say not for performance cars in general. The SL500 is a pig at nearly 4,000lbs and has way too much torque and horsepower, and as a result, oversized parts which aren't made for normal driving. As the "performance" of the car ramps up, so does the difficulty of handling the car. A possible exception to this rule is AWD cars.

>We can have separate race tracks for that.

We do have racetracks, and people want to drive regular cars fast on those racetracks.

I am all for shaming the auto manufacturers that get caught cheating. The regulations are aimed at reducing pollution and I think that is a good thing!

it's not about what the regulations AIM to do, it's what they achieve, and also if they could feasibly achieve something closer to their aim, which is what was suggested.

that's to say, if emission testing was done differently, maybe we would have caught these cars earlier. Now it seems likely every single non-electric car on the market is polluting more than advertised/allowed, which begs the question if there is a systemic issue here. I tend to agree that there is.

The 'systemic issue' is that conducting lab tests in cooperation with automakers is clearly an inadequate way of enforcing the regulations. That does not imply the regulations are faulty, it implies they need a more robust enforcement methodology.

yes, i see now i formulated that rather poorly. this is what I meant by systemic failure. the regulations themselves are not the issue.

What if the AIM was to create the incentive for electric cars manufacturing

As we've seen with Facebook, Equifax, and VW, shame has no lasting effects. It seems that even fines end up being slaps on the wrist.

The regulations might have a good aim, but when everybody is violating them it's probably safe to say something is wrong with them. I don't know if the problem is in the content or in the enforcement, but it exists.

I don't think criticism of regulations should replace shaming of auto manufacturers, but it might be more effective.

From a recent HN thread[0], it sounds like this issue is present across the auto industry:

> One way to reconcile these seemingly contradictory statements is the following. The actual programmer & designer of the software is the supplier of the engine control units (ECUs), in most cases Bosch. The software that they supply actually already contains these cheating routines. These routines are delivered as a part of the software packages, and can be turned on/off with some flags.

It is up to the auto company how to set these flags.


> The general assumption at the time when VW was put under a microscope for these issues is that everyone is cheating, VAG just got caught first.

Well, they do.

Here's a list of some car manufacturers who do or did emissions cheating:

- Volvo - Volkswagen - Renault - Jeep - Hyundai - Honda - Citroen - Fiat - Chrysler - Ford - General Motors - Toyota - Nissan - Mercedes - Mitsubishi

(Which pretty much sums up to "all car manufacturers")

This gets repeated on every thread on this topic but from what I've read is not true and heavily favors VW. VW was caught actively faking the test results by having the ECU detect it's being tested and change its settings. What every manufacturer has been caught doing is optimzing for the test and not getting the same results in real world conditions. That's a fault of the test and is perfectly legal. The only guilt of the manufacturers is lobbying to not upgrade the test.

IIRC Mercedes wasn't caught cheating they just got in trouble because they use the same engine control units which contain the routines that make cheating possible.

Mercedes was entering an allowable disclosed higher emission mode with an unallowable, undisclosed, entry condition.

In EPA terminology, a defeat device.

Mercedes was already ordered to recall affected models.

"everyone is cheating, VAG just got caught first."

Everyone was cheating in Europe, because European emissions rules were/are weak and you could get away with a defeat device provided it was for "engine protection".

The difference with VW was that they tried to pull the same thing in the USA, which had much stronger rules and a clear prohibition against defeat devices.

VAG cheated for years, got caught, then lied to regulators for a year, then got caught in the lie. That's how you REALLY piss off regulators (or law enforcement in general).

Also, yes, regulations should be better matched to real world performance.

Do you have any links to something I could read about the way the regulations are being implemented? It is relevant to my interests.

This incident only related to cars sold in Japan. The VW scandal was different.

I think it's great that they're falling on their faces... Hopefully it locks them into only 1 solution - all electric zero. Pumping gas sucks. Smelling gas sucks. Hearing gas burn sucks. Gas cars suck.

That is a ridiculous solution. Batteries are far from clean, what do you propose we do for disposal of these batteries? What about the environmental and human cost to greatly increasing production of materials required? All electric isn't the utopia people make it out to be.

> We found that battery electric cars generate half the emissions of the average comparable gasoline car, even when pollution from battery manufacturing is accounted for.


Yes, EV's have higher energy and emissions numbers embedded in their manufacturing, but that gets more than offset by the much lower emissions of fuel. Breakeven is the first 6-18 months, depending on battery size and emissions makeup of your local grid. Even in "coal-powered car" country.

> EVs are powered by electricity, which is generally a cleaner energy source than gasoline. Battery electric cars make up for their higher manufacturing emissions within eighteen months of driving — shorter range models can offset the extra emissions within 6 months — and continue to outperform gasoline cars until the end of their lives.

As for waste, there is battery recycling, but the more interesting part is second-life storage use. Battery may not have the energy density required for transportation, but you can still use them fine as a cheap powerwall-like product. Cheap because the cost of the batteries was already paid for by the first owner of the car.

So because batteries are not perfectly clean means we should keep using gasoline which is worse?

Subjectively worse, we don't have enough battery powered cars to compare the timelines with gasoline powered. They contribute to more waste, regardless. Battery packs need replacing 5 - 10 years, are not easily recycled or disposed of and their manufacture and mining of materials still isn't a net positive in terms of pollution. There isn't a one size fits all solution. Going all electric trades one issue for another, cleaner air for more polluted rivers, cities around mines and factories and battery disposal and recycle sites. I personally have no problem with electric but I dislike hearing people tout it as the solution to all our vehicle woes. It isn't.

That's the answer... shift the emissions burden to power companies and battery disposal companies! But I'll feel good driving my "zero environmental impact" electric car around!

I think you meant this as snark, but I agree with it non-ironically. Shifting the emissions burden to centralized power plants makes later reductions simpler to achieve as power generation migrates to cleaner technologies. It's easier to determine how dirty the energy is when it's being generated at N places rather than 1e6(N) places. Emissions per watt decrease as new solar/wind capacity is added to the grid. You have full control over your energy source with an electric car (i.e. you can switch to solar now if you'd like and charge from that).

Actually yes, despite your tone. It's significantly easier to control emissions and implement emission control systems at the production side because economies of scale means that for the same cost of making $number of cars cleaner you can make 1 power plant significantly cleaner using far more expensive technologies and still be more cost efficient.

In the auto industry its generally understood that the emissions limits are so low, especially for diesel, that its physically impossible to meet them in some situations like cold engine + cold weather and when driving up grade.

Emissions regulations have made great progress in cleaning up auto emissions, but now politicians keep squeezing the numbers without realizing some of the limits are becoming impossible. So everyone starts cheating, because otherwise we would have no combustion powered cars.

Insane irony is that ships and even some buildings burn horrifically impure "bunker fuel" or wood, and with car emission limits so low these sources are the vast majority of harmful emissions. Its like cities in Cali forcing citizens to make drastic changes to conserve water, when over 90% is used by farmers.

> especially for diesel

There's another way of looking at this, and that's that maybe diesel just isn't clean enough for consumer cars?

maybe, but its cleaner in other ways. It burns hotter and more efficiently, creating less CO2 and some other byproducts by a decent margin. It produces more SO due to fuel sulfur content, and more NOX which is produced because of high temperatures. So its a trade-off.

But at this point the limits are just politicians pulling levers. There's not enough science involved to debate which pollutants we care about and what the lowest realistic limits are.

Forbes says that it would have cost an extra $430 per vehicle for VW to meet the NOx requirements fairly. That's hardly the same as regulations resulting in no combustion-powered cars.


>Its like cities in Cali forcing citizens to make drastic changes to conserve water, when over 90% is used by farmers

yes, this is a common process by which private costs are offloaded onto the public so as to increase the private benefits of exploitation of public goods.

it's legalized theft.

Completely agree that what is lacking here is the regulations. I have tried to argue (often unsuccessfully) that VW did not commit fraud, that the numbers reported by the tests were 100% accurate. The problem being that the way the car is tested is not even remotely similar to how people drive. The clear fix to me is to simply give normal people free leases for a short period of time and generate something similar to http://www.fuelly.com/ . The only way to cheat that is to convince people to drive more efficiently, seems like a win.

How is embedding software that changes the way the engine behaves during tests versus during actual usage not fraud?

The numbers may have been accurate _for an engine with the 'limit emissions' mode enabled_ but it's still fraudulent if you do not make that explicitly clear.

If you use the car the way tests use the car, the performance will be the same. The trick is you can't complete your daily driving needs by driving the way the test does.

The only way to drive in a way that would have replicated the results of the VAG tests would be to not touch the steering wheel. At all.

For VAG this wasn't a case of the test not being quite the same as regular driving, it was VAG deliberately detecting a tests and altering the way the engine worked to avoid failing.

The intent was to cheat the tests by secretly tailoring the behavior to the tests, even though they knew that real-world driving would be different and that the tests are supposed to indicate real-world emissions.

Intent matters. You can’t set out to cheat a test, succeed, and then say that it’s legal because you succeeded, any more than you can steal from a store by hiding an item under your coat and then say that it’s legal because you disguised it well enough to fool the guard.

If you turned the steering wheel the test would fail, you can't drive without turning the steering wheel.

It would also turn off if the back wheels were spinning.

That's the whole point: you cannot use the car the same way as during the tests, even if you exactly match speed, acceleration, distance, friction, temperature and any other variable. The software that limits emissions is disabled during "normal" operation, ie. driving.

I will attempt to make an analogy that tries to shine light on the angle I'm coming from. When we interview for software engineers, many companies place much of the emphasis on algorithms as opposed to other hard-to-test skills. Now, you have a candidate that aces any algorithm question put in front of them yet completely sucks at software engineering. Did that person commit fraud? It seems, to me, that the test is what is broken.

We all clearly understand the angle you're coming from, but you haven't been able to make the case it isn't fraud. Because it is textbook fraud.

If you think the test is bullshit, then you make that case separately. You don't lie about how you're passing it and then when caught say 'but it's the test's fault!'.

Emissions tests are different from fuel efficiency tests, and in fact are often at odds with each other. A more efficient car doesn’t necessarily have better emissions, and there’s no easy way for a typical driver to measure their emissions.

Thank you for the extra insight, I may have had the wrong conclusion but I think the logic still follows. From this system I outlined, the manufacturers would be forced to alter designs to get the desired emissions during normal driving conditions. This is the whole spirit of the regulation right?

My point is that the system you outline doesn’t work, because ordinary drivers can’t check their own emissions. It would be fine for fuel consumption since that’s easily measured, but emissions requires special equipment and careful testing.

NOx emissions specifically are a problem because the engines are run very efficiently at high burning temperatures with complete combustion.

The VW stuff wasn't simply about fuel efficiency it was about emissions output. I think this was the video where a guy basically reverse engineered the system. (Starts around 40m in) https://media.ccc.de/v/32c3-7331-the_exhaust_emissions_scand...

I'd encourage you to watch that, and still see if you say that it wasn't fraud.

VM gamed the test in order to produce results that were not achievable by consumers, with the intent to deceive people.

I will watch that video later and get back to you. I understand what they did. I'm not saying that what they did was right, simply that it wasn't fraud. My point being that if the test is already deceptive for true vehicle emissions (laboratory settings, very particular configurations, etc.) then the test itself is the problem. If we flip the script and say that a car fails the emissions testing but is in reality the cleanest car, is it not "fraudulent" to label this car as dirty?

Intent matters. Your example of flipping the script would only be comparable if the manufacturer engineered the car to fail the test.

VW built their car to produce a specific test result to get buyers to buy the car. It fits the description at the link below all too perfectly.


Without knowing the general driving population's driving habits, it's difficult to estimate how effective driving will improve fuel efficiency and/or reduce pollution. Sure there are known bad driving habits like sudden accerlation/sudden stops will increase fuel consumption, but what about road sign design, speed limits, traffic light designs and Electronic Toll Collection systems? Those are all factors of fuel efficiency.

Also, the government regulation on emission is going to get tighter and tighter due to incentives to push for EV adoption. So arguing emission test method is a moot point.

Regulations are obeyed and enforced by humans. There’s a general expectation that common sense is applied, which is hard to define but “you know it when you see it”.

“The engine will fundamentally perform differently when you’re testing it because it will detect the test” is not common sense and is not a reasonable interpretation of how to comply with the rules.

I agree that a better method would be to attack sensors to a random (and ideally, concealed from OEM) sample of cars and charge automakers per unit of detected emissions. That would force them to optimize for real world driving, which is much harder to game.


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact