Oral evidence: Volkswagen Group emissions violations, HC 495
Oral evidence, 25 January 2016.
Witnesses
Paul Willis, Managing Director, Volkswagen UK, and Oliver Schmidt, Engineer, Volkswagen AG, gave evidence.
Chair
Good afternoon and welcome to the Transport Select Committee. Would you give us your name and position in the organisation, please?
Paul Willis
I am Paul Willis, managing director of Volkswagen UK.
Oliver Schmidt
I am Oliver Schmidt. I am an engineer by trade. I work with Volkswagen AG in Germany. My position is running the office for group powertrain development, but when the software issue with the diesels occurred I was asked to support diesel development in Wolfsburg and to be one of the main contacts for agencies worldwide.
Chair
Thank you very much. Mr Willis, thank you for returning to the Committee. We asked you a number of questions last time; some you were able to answer and some you were unable to answer. There have been a number of developments since that time. Mr Schmidt, we are very pleased that you are here as a technical expert. There were a number of questions put last time that Mr Willis was unable to answer because of lack of technical expertise. That is the reason you are here today, and we thank both of you very much for coming. Mr Willis, you have previously outlined to the Committee Volkswagen’s planned roll-out of vehicle modifications in the UK. Have all the affected owners received letters informing them about what is happening?
Paul Willis
That is my understanding. We have sent out two sets of letters. We sent the first set of letters to customers in early October to say that their vehicles were affected. We sent a further letter before Christmas with more information and to tell them that we were developing the fixes. We gave them approximate timelines and said that we would come back to them at the appropriate moment. Also, as you remember, we put a VIN recall on the internet so that people could self-serve, and we have used other methods to communicate with customers via the internet. We have done everything, but we must continue communication with customers, because we know from our research that not everybody fully understands the issue. For example, of the 1.2 million customers who were affected we had communication with circa 40,000 people, which is less than 4%; 40,000 people either wrote to us to ask more questions or, indeed, they—
Chair
Mr Willis, we are aware of the extent of the problem. We are trying to find out exactly what is being done. What kind of feedback have you had from the letters that have gone out?
Paul Willis
We have had very variable feedback right across the spectrum. In the main, the feedback has been reasonably positive.
Chair
Tell me where it has not been positive.
Paul Willis
We have had some customers who do not accept what we say. We have some customers who have had several communications with us, and we are discussing and trying to deal with them as best we can.
Chair
What are the issues that they have brought up?
Paul Willis
The issues that people mainly bring up are the fact that they are asking about the car in real-world driving and they are asking about compensation.
Chair
What kind of help are you giving the dealerships, who must have been affected by what has happened?
Paul Willis
We are giving an enormous amount of help to dealerships. Obviously we have all the communications that I have spoken to you about. We have helplines. We are very open with our information. I met with every dealer council personally within two weeks of the crisis breaking, so I have been communicating with dealers. We have also supported dealers financially with stocking, because the circa 4,000 vehicles that we spoke about last time, which we have still not sold, are on our stocking plans. On certain brands we have given the dealers more money to help their sales, so I am very confident that we have been very fair with dealers.
Chair
Are the dealers satisfied with the action you are taking?
Paul Willis
In the main, the dealers are happy. I can report that we had some sales issues in October, which we expected. We had some order issues from customers in November, but our retail orders—
Chair
What were the order issues and sales issues?
Paul Willis
The sales issues were below where we expected, but the most important point is that, through our communication and through explaining to people, in December we had more retail orders than we had last year. In January we also had more retail orders than last year. My conclusion is that, overall, in the main, people understand what we are saying, but I cannot hide from the fact that there are some people who do not accept what we say, and we are in communication with them.
Chair
How are you dealing with the negative feedback that you have been getting?
Paul Willis
We are writing to people and talking to them as and when they discuss it with us.
Will Quince
You said 40,000 out of just over 1 million. That seems like a tiny percentage that you have written to. I understand that there must be people who have sold on their vehicles since, but is there no way you can work with the DVLA to ensure that more customers are made aware of the issue?
Paul Willis
Sorry; let me make myself very clear. We have communicated with 1.2 million people twice by letter. We have had communication back from 40,000 of that 1.2 million, which is a tiny amount. You are right. Forty thousand people, less than 4%, have come back to us with some questions. I have to say that most of those people said, “Okay, we understand. Tell us when the fix is coming.” A number of those people are still not happy with what we say. I have to be very open about that.
Will Quince
Did you have contact details for the 1.2 million or was that provided by the DVLA?
Paul Willis
Obviously from the very outset we have been in contact with the DVSA and the DVLA. They gave us all the information and we immediately wrote to them and communicated with them.
Chair
Mr Schmidt, what can you tell us about the software update? What difference has it made to NOx levels during emissions tests compared with vehicles that have not had it?
Oliver Schmidt
The amount of NOx levels stays the same. We were within the legal limits for NOx levels before, and afterwards the NOx levels were within the legal limits as well.
Chair
Are you telling me that the software update is not making any difference? Is that what you are saying?
Oliver Schmidt
The software update removes the software that recognised the drive trace, but during an emissions test it has no impact.
Chair
It has made no difference.
Oliver Schmidt
No.
Chair
Is that what you are saying?
Oliver Schmidt
Yes.
Chair
That sounds most remarkable.
Paul Willis
Mr Schmidt is absolutely right. In the test there is no difference in the NOx emissions; with little changes in technology he will tell you that in real-world driving there is a small improvement, but it is small.
Oliver Schmidt
I understood that you specifically asked about the test.
Chair
What kind of testing have you done with the software modifications?
Oliver Schmidt
With the software modification we test the software in the new European driving cycle, because that is the legal requirement. We test it in the new European driving cycle under warm conditions. We test it in the WLTC or the WLTP, which is the new upcoming emission test in a few years, and we test it in a special cycle developed by the German ADAC, which is comparable to your RAC in the UK; it is a roadside assistance and consumer protection agency. They developed a special German highway test. We test the vehicles in those four tests. As Mr Willis pointed out, because of advances in technology and engineering we see decreases in the values that are not in the European driving cycle. We see decreases because we can now utilise engineering knowledge that has been gained over the last 10 years.
Chair
Are you seriously saying to us, Mr Schmidt, that the software devices that have now been put in to test the engines are not showing any difference in NOx emissions?
Oliver Schmidt
No; as I said, they do not show any differences in the new European driving cycle, where we are below 100 mg per kilometre, which was the legal limit for Euro 5 vehicles. We see improvements in the other areas that vary from engine type to engine type, because of the advances we have gained in engineering over the last 10 years.
Chair
What type of real-world testing have you done?
Oliver Schmidt
Real-world driving testing was never part of Euro 5.
Chair
I am asking you what real-world driving testing you have done.
Oliver Schmidt
We have not performed any real-world driving testing with Euro 5 vehicles.
Chair
I thought you just said that you had done small amounts of real-world driving testing.
Oliver Schmidt
No. We got the drive traces for those tests and we programmed them into our rolling roads to have repeatability for the tests. If you perform a real-world driving test, the result varies from test to test because you always have different driving conditions; you have different ambient temperatures and the driver has a huge impact, so we programme those tests into our rolling roads and perform the tests in our labs.
Chair
How has the software device affected fuel consumption?
Oliver Schmidt
From all engineering measurements I have seen so far—we are talking miles per gallon—there is no decrease in miles per gallon.
Paul Willis
If I can be very specific—
Chair
Let me ask Mr Schmidt. You said “from all engineering measurements.” What does that mean? What kind of testing have you done to find out if it had an impact on fuel consumption?
Oliver Schmidt
Fuel economy for those vehicles is determined in the new European driving cycle. We have performed those tests and we have not seen an increase in CO2 or a decrease in mpg in those tests.
Paul Willis
May I just add something?
Chair
Let Mr Schmidt finish. What are you telling me? You have not done any real-world driving test. You have been doing something and you do not think it affects fuel usage in real-world driving, although you have not actually tested for that. Is that what you are saying?
Oliver Schmidt
The problem is that for Euro 5 vehicles there is no described method of real-world driving testing.
Chair
I didn’t ask you a question about that. That was not my question. My question is: how has this software alteration affected fuel usage?
Paul Willis
It has not.
Oliver Schmidt
It has not—
Chair
Can Mr Schmidt answer?
Oliver Schmidt
As I pointed out, we performed the tests I mentioned earlier. In those tests we have not seen an increase in CO2 values or a decrease in mpgs. We performed the four tests that I mentioned earlier and we have seen no differences there.
Chair
You are confident that the additional software will not affect fuel consumption for motorists.
Paul Willis
Yes—
Oliver Schmidt
It is not additional software. It is software. We remove what is currently in the ECU. We remove the switch logic that we recognise as the drive trace and then we have software. We do not add software; it is always one packet of software.
Chair
You have removed what have been defined as defeat devices.
Oliver Schmidt
Not in Europe.
Chair
You have not removed them.
Oliver Schmidt
No, this software is not defined as a defeat device in Europe.
Chair
You are incorrect. The German transport authority has been very clear in its ruling that what you were doing was not legal in Europe. Are you telling me that you have not removed that?
Oliver Schmidt
The legal assessment of the Volkswagen Group—
Chair
I am not talking about the assessment of the Volkswagen Group. Are you telling me that you have not removed the device that was described as a “defeat device” and seen as improper, if not illegal, by the German transport authority?
Oliver Schmidt
No.
Chair
Are you telling me that you have not removed it?
Oliver Schmidt
No. I am telling you that we have removed it, but I said that in the understanding of the Volkswagen Group it is not a defeat device.
Chair
I did not ask you that. I asked you what you were actually doing. You are not here to tell us what the Volkswagen Group say about a question I have not asked you. I am trying to establish exactly what has happened. Mr Willis, you wanted to answer?
Paul Willis
Thank you very much indeed. The process is that the cars are tested in a great level of detail, as Mr Schmidt told you. We then go to the KBA and have sign-off from the KBA. After what our company has been through it is very important to re-establish trust, and one of the ways we will re-establish trust is by having sign-off from the KBA, which is part of the process. If I can just report, the first vehicle—
Chair
Let me stop you there. Can you now take me through the process in Volkswagen for developing and approving the software? What was the process that was followed?
Paul Willis
Oliver can talk about the process and I will talk about the investigation.
Oliver Schmidt
What we established in Volkswagen as a result of this is that we now have a “four-eye” principle on software development. The base code is now checked by quality assurance, which was not done before. All software was just developed within engineering, but now, before the software is encrypted at the supplier, our quality assurance checks the software under the four-eye principle. That is what has been changed.
Chair
What happened before? Was there any quality assurance?
Oliver Schmidt
No, there was none.
Chair
There was nothing before.
Oliver Schmidt
No.
Chair
How has it developed? It just happened, did it?
Oliver Schmidt
It was developed within engineering and then put into production.
Chair
With no process of quality assurance.
Oliver Schmidt
Of course there was always a process of quality assurance, but there was no process where the quality assurance looked into the source code of the onboard computers.
Chair
What happens now?
Oliver Schmidt
The new step, which has now been implemented, is that quality assurance within Volkswagen looks into the source code of the ECU before it is encrypted.
Chair
Mr Willis, do you want to add to that?
Paul Willis
Yes. There was a very interesting description in the press release of 10 December, to which I can refer you, from our supervisory board chairman, Mr Pötsch, where he talked about some of the adjustments—the new way of doing these things. He said that development would be in strict “accordance with the 4-eyes principle” with better use of IT to check processes to ensure they are properly adhered to—in other words checking work flows— and they would ensure that “future emissions tests will be verified…independently” and “randomly selected real-life tests to assess emissions behaviour on the road will be introduced.” It is clear that those processes were not actually taking place previously in engineering, and that is what we apologise for. We really have to look at those processes, but we also have to find out from the Jones Day investigation precisely what happened and why. We are still waiting to receive the report from Jones Day.
Chair
Who actually signed off the software upgrade? Mr Schmidt, can you tell us that?
Oliver Schmidt
Do you mean in the past?
Chair
No, the software upgrade.
Oliver Schmidt
As Mr Willis pointed out, and as I mentioned earlier, it is now a four-eye principle between engineering and quality assurance.
Chair
Who signed it off? Who is responsible?
Oliver Schmidt
The head of engineering and the head of quality assurance.
Paul Willis
Two people. Four eyes means that nobody can do anything on their own. The engineers say, “Here is the software,” and they present it. Then they have to present it to someone else who is completely outside the department, in quality assurance, to check that it is absolutely in line with all regulations.
Robert Flello
In relation to the people who are now signing off on this software, how much code are they looking at? Are we talking about five or six lines of coding or 1,000 lines of coding?
Oliver Schmidt
In total, in an ECU, if you print the software in the affected engine, it is about 20,000 pages.
Robert Flello
Those two sets of eyes are looking through 20,000 pages, looking for something anomalous.
Oliver Schmidt
Yes. Of course, they were trained and were shown what was in the software before. They were trained about what to look for.
Paul Willis
Mr Flello, if you really want to find out the detail of who signs off what post the Jones Day report, I am more than happy to invite you, or members of this Committee, to look and check for yourselves to see how it is signed off. That invitation is open to you.
Robert Flello
That sounds marvellous, but I am more interested at the moment in the two sets of eyes. They are the head of engineering and the head of quality assurance.
Oliver Schmidt
It is not the head of engineering and the head of quality assurance; it is the people working for them. Mrs Ellman asked me who signs off in the end, and of course in the end the responsible board member signs off.
Robert Flello
The two responsible board members are signing off on a piece of work that their subordinates, in their teams, do on 20,000 or so lines of coding. How does that differ from before?
Oliver Schmidt
There was no four-eye principle before. The software was developed in engineering and put into the cars and into production. The only checks were driveability checks.
Robert Flello
Coming back to the four eyes; it is not four eyes but two teams of eyes that are looking through the code.
Oliver Schmidt
Yes.
Robert Flello
Nobody looked through the code before; you just took whatever the engineers came up with in 20,000 lines of code and that was fine. Nobody looked at it or studied it. You were just reliant on the engineers.
Oliver Schmidt
As Mr Willis pointed out, we will have to wait for the final Jones Day report, but to my knowledge so far that was the process before.
Huw Merriman
The defeat device, as many call it but you do not, enhanced the NOx results. You then take out the defeat device and due to new technology, from the last 10 years, you can then net off any of that NOx impact.
Oliver Schmidt
Correct.
Huw Merriman
If that is the case, and it has been there for 10 years, why did you need to put in the defeat device in the first place? The technology was already there.
Oliver Schmidt
Again, I have to refer to the Jones Day report.
Paul Willis
We do not know.
Oliver Schmidt
I do not know who put it in or why. During the time it was developed I was not in powertrain development. I was in group production. I cannot tell you who put it in or why. I have to refer you to the report, but what I can tell you is that there are versions of this engine on the street today that are not affected; they do not have the software in them and they meet emissions.
Huw Merriman
Are you saying it was all a red herring—a completely irrelevant and redundant bit of kit?
Oliver Schmidt
No, I did not say that. I said I cannot tell you who put it in, but there are vehicles on the street that meet the emissions without the software.
Paul Willis
You make a very good point though, and I want to re-emphasise that point. The emissions were affected in testing; in the real world and for real customers there is no effect.
Will Quince
I would like to pick up on that point. It just does not seem to make sense. There is a relatively easy fix, which Mr Schmidt tells us you have done via a software upgrade.
Paul Willis
Correct.
Will Quince
Coming back to Mr Merriman’s point, why do you believe that somebody—two rogue engineers or lots of engineers, who knows—would do this if it was such an easy fix?
Oliver Schmidt
I cannot tell you. I have no knowledge about who did it or why they did it. I have to refer back to the Jones Day report that is supposed to be published, at least in part, during our shareholders’ meeting in April. I hope we will then get those answers. I would love to get the answer myself.
Chair
Let us keep to the questions on testing. We will come to the report, because it is a very important area. What testing does Volkswagen carry out now to ensure conformity of production? You are supposed to be testing to ensure that correct standards are maintained. What is happening now?
Paul Willis
What is happening now is that one month after the discussion on NOx, and shortly after I had the pleasure of coming to see you for the first time, we had an announcement that there might be some problems with CO2. The reason is that some of the people in—
Chair
That is not my question. I am asking you about conformity of production. What are you doing in relation to that?
Paul Willis
That is what I am getting to. It is very simple now. We are working with all the different agencies. If any cars currently look as if they are slightly in the red area—which is, by the way, a normal process under conformity of production—some of them go through a testing regime right now with the relevant authorities; for example, we are waiting for the VCA to come back to us so that we can start discussions about what their points are. We say that on the CoP, it is this. They have gone in and witnessed it in Mladá Boleslav in the case of Skoda, and now the discussion starts.
Chair
But there are two separate things. As a company, you are responsible for establishing a conformity of production process to ensure that standards are maintained.
Paul Willis
Yes.
Chair
The type approval agencies, such as the VCA here in the UK, have their own responsibilities to see that you are doing it properly, but you have an initial responsibility. What is changing now in relation to how you do that?
Oliver Schmidt
The software in question was effective every time a car was put on the new European driving cycle. It could not have been detected under conformity of production, but, as announced on 10 December, we will now perform a certain number of conformity of production tests as real-world driving tests. This has not been done before, but it has been announced. Testing of the current product will now not only be done on the dyno but on a certain percentage on the street with a PEMS device.
Chair
Will you be publishing the results of those tests?
Oliver Schmidt
I am not aware whether we are going to publish those results, but every manufacturer is mandated or forced by the new European emission regulations to publish their real-world measurement results obtained during the monitoring phase and then later during the final phase on their website. Yes, we will be publishing, because it is the law. The first vehicles where we publish those numbers will start in calendar week 11.
Chair
Is that right, Mr Willis? Will the company be publishing the results of the real-world driving testing?
Paul Willis
If you are specifically referring to the conformity of production on CO2, we are in discussion with the authorities and of course we will publish them, yes.
Chair
It is not just CO2—real-world driving emissions.
Oliver Schmidt
Real-world NOx emissions, yes.
Chair
They will all be published.
Oliver Schmidt
Yes. It is mandated by law that you publish those values. It is part of the European regulation.
Robert Flello
I want to clarify one thing about the software. How regularly over the last 10 years has the software on the ECU been updated? Quite often when a car goes in for routine maintenance, an upgrade is popped on to the ECU or there is a reconfiguration. Roughly how many times over the last 10 years would that have happened?
Oliver Schmidt
Do you mean on one specific vehicle?
Robert Flello
Yes.
Oliver Schmidt
I do not have an answer here.
Robert Flello
Once? Ten times?
Oliver Schmidt
No, not 10 times; maybe two or three times. It depends on the type of vehicle and it always depends on whether you detect problems in the field through your CoP testing or through any customer complaints. Of course, once you detect such a problem you have to react and update your software.
Robert Flello
The software engineers who are working on upgrades or updates generally would not have noticed that there was a strange bit of code in the—
Oliver Schmidt
They were the same. There is one set of engineers who work on the software code for ECUs. It is not several sets of engineers.
Robert Flello
It is the same group of engineers who have been working on the code for the last 10 years.
Oliver Schmidt
Yes.
Graham Stringer
I am puzzled, Mr Willis. The first statement you made to the Committee on 12 October was an apology.
Paul Willis
Yes.
Graham Stringer
What were you apologising for?
Paul Willis
I was apologising because what we did was to put a car on a dynamo in a laboratory and the car recognised it was on the dyno, and as a result it changed the NOx characteristics. That was inappropriate, and the KBA said that was inappropriate. What we needed to make sure was that we worked within the appropriate boundaries of the KBA. That is the simple reason.
Graham Stringer
But you maintain that that recognition of the testing conditions—that the car was under test—was not a defeat device under European law.
Paul Willis
That is correct; yes.
Graham Stringer
Can you expand on that? The notes I have before me say that a device that recognises testing conditions is not compliant with European regulations.
Paul Willis
European regulations say that it has to be part of the emission control system. European regulations also say that it has to be under conditions that may reasonably be expected to be encountered in normal vehicle operation. Our position is that that is absolutely not the case.
Graham Stringer
You are saying that even though your device recognised that the car was being tested, according to your legal advice, it is still compliant with European regulations.
Paul Willis
That is our position, yes.
Graham Stringer
How much money are you spending on lawyers at the moment against possible action by different authorities?
Paul Willis
I have no idea about that.
Graham Stringer
Why not?
Paul Willis
Why should I know how much we are spending on lawyers? I don’t know. Across the world I don’t know. I am responsible for Volkswagen Group United Kingdom. I don’t speak to lawyers outside the United Kingdom.
Graham Stringer
How much are you spending in this country?
Paul Willis
I don’t know. I really genuinely don’t know.
Graham Stringer
Isn’t that odd for a person in your position?
Paul Willis
I don’t think so. I speak to my colleagues maybe once a week. No, I don’t think so. I am actually concentrating, Mr Stringer, on trying to satisfy customers, regain trust and get these cars fixed. That is my concentration at the moment.
Graham Stringer
But I would have thought part of your responsibility to shareholders would be about costs.
Paul Willis
Of course.
Graham Stringer
If you are potentially subject to legal action by a number of regulators, I would have thought you would want to know what your potential liability was, and therefore would be hiring in lawyers. I just find your answer incredible.
Paul Willis
I can talk you through the financial results of Volkswagen United Kingdom, but I do not think that is appropriate here.
Chair
No. We want the answer to Mr Stringer’s question.
Graham Stringer
If you are telling the Committee, when you are potentially liable to action by a number of different regulators and individuals, that you do not know how much your business is spending on lawyers, I find it incredible from somebody in the top job. But if you are telling the Committee that is the case—
Paul Willis
That is the case, yes. What I can tell you is that if you go back to the press releases you can see that our company has provisioned—I cannot remember the exact number—something over €6 billion.
Chair
Mr Willis, we will be asking you more questions about that specifically. Mr Stringer was asking a slightly different question.
Stewart Malcolm McDonald
Could you provide the Committee with an update on the Jones Day investigation?
Paul Willis
I can tell you three things about the Jones Day investigation, which are along the lines of what I heard on 10 December. There is an enormous amount of information to go through: over 100 terabytes, which I understand is the equivalent of 50 million books. I can understand that that will take a very long time. We have employed Jones Day and other advisers to help us, and 450 people are currently trawling through people’s devices; over 1,500 employees’ devices have been taken to be looked at. We know from what we have found so far—answering Mrs Ellman’s earlier question—that it is clear that some of the work processes needed to be improved. We are waiting to hear what the rest of the report has to say. The supervisory board has committed that the next formal update will be when we publish our 2015 results. I think the date is around 21 April.
Stewart Malcolm McDonald
When you came to the Committee in October, you said that you thought this was down to the work of a few rogue engineers. You have just outlined that 1,500 devices are being looked at by 450 people—the equivalent of 50 million books. Has anyone senior in Volkswagen UK been suspended?
Paul Willis
Not at all; absolutely not. I want to come back on that question. Just because someone has said, or I have said, that there were one or two people—that is conjecture and we do not know—it does not mean that we are not trawling through all the information. We are looking through all the information because we have to find out what went wrong. It is very widespread. Coming to the UK, I can categorically say that neither I nor any of my employees had any idea of what was going on. I repeat what I said at the last Committee for the avoidance of any doubt: I did not know what a defeat device was before this, though I now know what a defeat device is, and neither did my employees.
Stewart Malcolm McDonald
You cannot say that until the Jones Day inquiry concludes, can you?
Chair
Mr Willis, you did say that neither you nor any of your employees knew. How could you possibly know that until the inquiry is completed, unless somebody is not allowing the inquiry to access all the information it needs?
Paul Willis
Because I am a sales, import and finance company. That is why I can say it. It is very clear. We have nothing to do with engine development. I am really clear on that, ladies and gentlemen.
Stewart Malcolm McDonald
How would it be possible for a reputable company such as Volkswagen to be engulfed in a scandal this big and for it only to be the work of a handful of engineers? Do you honestly think that the public at large—I do not just mean Volkswagen consumers—buy that, Mr Willis? I will be honest with you; I don’t, and I don’t think this Committee does.
Paul Willis
I will give you my opinion. My opinion is that I think it was very few people. We will have to wait until the Jones Day report comes out. Like you, I am bemused about how something like this could happen. I am really bemused. I have been associated with this company for over 15 years. Let’s wait and we will find out when the Jones Day report comes out.
Stewart Malcolm McDonald
Do you think that Volkswagen is an ethical company?
Paul Willis
Yes, I do.
Stewart Malcolm McDonald
Do you think it knows the difference between right and wrong?
Paul Willis
Yes, I most certainly do.
Stewart Malcolm McDonald
To follow up on Mr Stringer’s points, you seemed to dance on the head of legal jargon as to whether or not it was a defeat device and whether it does or does not break European law. Don’t you think that what people see is that the Volkswagen Group, essentially, is “at it,” as we say in Scotland? You have perversed environmental regulations. You have treated European customers with disdain. You have treated regulators like bureaucrats and people are quite fed up, and not just with the actions of Volkswagen. We now see Renault potentially being in a similar situation. You talk about your company being ethical—
Paul Willis
Sorry; who is in a similar situation?
Stewart Malcolm McDonald
Renault is potentially in a similar situation. You talk about your company being ethical and knowing right from wrong, but isn’t your entire industry dogged by this kind of problem, where you have tried to get around regulations all over the world?
Paul Willis
No, I absolutely refute what you are saying.
Stewart Malcolm McDonald
Do you really?
Paul Willis
Yes, I absolutely do. Volkswagen is an upstanding company. It cares about its customers. It invests billions, more money than any other car company, in technology. It is important that we get to the bottom of this. By the way, I do not agree with what you say—that we treat people and customers with disdain. I am very clear that none of this affected customers in any way. Mr Schmidt says, and I believe him and I believe the KBA, that there will be no impact on fuel consumption, but I would like to understand completely what has gone on. Absolutely. With regard to the industry, I think it is very important for the United Kingdom and Europe.
Stewart Malcolm McDonald
It needs to clean its act up though, don’t you think?
Paul Willis
What I think is that we all have an obligation to become transparent, with testing in the real world. We need to make people more aware, because that is really the issue.
Chair
Mr Willis, when you were asked who had been suspended, the question was about senior employees. Have any employees been suspended?
Paul Willis
Yes. Nine employees have been suspended, but I don’t know who they are. Under German privacy laws that is the case.
Chair
What level are they?
Paul Willis
Again, I am sorry to report that I don’t know specifically, but I understand they are of varying levels. We will find all this out in the Jones Day report. We are not trying to hide anything from you, please believe that.
Stewart Malcolm McDonald
On that final point about not trying to hide anything, can you answer two things for me? Who will see the report before it is published, and will they have the opportunity to edit it before it is published?
Paul Willis
Let’s take the first one first. The report will be reported on at our results. If it is not finished, we will report on it again. I am not sure that the entire report will be given out where there is competitively sensitive information. I must draw your attention to that. I know Mr Pötsch very well, the chairman of the supervisory board. He is the gentleman who talked about the changes our company is making. I have known this gentleman for a very long time.
Stewart Malcolm McDonald
Will he be able to edit the report before publication?
Paul Willis
No, not at all. He is a very—
Stewart Malcolm McDonald
He will not be able to edit the report.
Paul Willis
I will come to that. Can I just finish? I am very convinced that we will be open and transparent, because our future depends on it. I do not believe anyone would edit the report.
Stewart Malcolm McDonald
But will anyone have the right to edit that report? I enjoyed your waxing lyrical there, Mr Willis, but will anyone have a right to edit that report?
Paul Willis
I would find it implausible, if you have independent legal—
Stewart Malcolm McDonald
So you don’t know.
Paul Willis
I find it implausible that if you employed independent lawyers you would edit the report.
Stewart Malcolm McDonald
But you cannot guarantee that it will be totally independent, because you do not know whether or not Volkswagen will have the ability to edit the report before the public see it.
Paul Willis
I don’t know. I will not be there, but I think it is implausible.
Chair
Mr Schmidt, do you think it is possible that the defeat device could have been installed by a group of rogue engineers without senior management knowing anything about it?
Oliver Schmidt
I have to refer you to the Jones Day report.
Chair
No. I am asking you, in your opinion, could that possibly happen from your knowledge of how things work?
Oliver Schmidt
Yes.
Chair
It is possible.
Oliver Schmidt
As I described earlier, we did not have the four-eye principle in the past; engineers coded it, so of course if there was nobody supervising or checking what they did, it could have been possible.
Chair
What could be their motivation?
Oliver Schmidt
This is what got published—
Chair
From your knowledge and experience, you are telling us that it is possible that such a thing could have happened. What could their motivation be in doing such a thing?
Oliver Schmidt
As published on 10 December, the motivation was to meet American emission regulations.
Chair
It could be that a small group of engineers, without anyone seeing or knowing anything about it, took a decision of that nature to do with that strategic view. That is what you are telling us.
Oliver Schmidt
To my knowledge and understanding that could have been possible in the past, yes.
Will Quince
On that exact point, what possible incentive could there be for engineers to take that unilateral decision, which could have such far-reaching impacts, as we have seen it has, on the business as a whole?
Oliver Schmidt
I could only speculate.
Will Quince
You are an engineer, so could you speculate?
Oliver Schmidt
I can keep costs down. That would be my speculation.
Will Quince
I understand that, but how can that possibly be a motivation or an incentive for the engineers? It would be a motivation or an incentive for management but not for engineers.
Chair
It’s a financial issue, isn’t it? Do you think it is likely that a group of engineers would take such a major decision on a financial basis that they were not employed to consider?
Oliver Schmidt
As I said before, during the time this happened I was not in powertrain engineering, so I was not directly involved in the processes. I cannot tell you who did it or why they did it. I can only speculate. The only thing that comes to my mind is that people wanted to meet American emission laws with the technology available at the time.
Iain Stewart
Mr Willis, I would like to go back to your comment in answer to Mr McDonald’s question, that no UK employees have been suspended or otherwise implicated in the investigation.
Paul Willis
That is correct.
Iain Stewart
Given, therefore, that the UK team are completely blameless, have any members of staff, many of whom are my constituents, suffered financial detriment either through redundancy or loss of overtime and bonuses or anything like that?
Paul Willis
Not as yet, but we have around 900 people in Milton Keynes, and when the crisis first broke I had a very worried workforce, for the obvious reasons. As news broke, I stood together with them. We have one big hall at the front of our building. It is the only hall that we can get 900 people in, and almost on the stump I would speak to them. I would then open the floor to questions, which is when I found how deeply concerned people were. It was coming up to Christmas. People were thinking about Christmas presents and their future employment. Of course we spoke about that a great deal. I spoke to people in Germany about it. At this stage I do not intend to make any redundancies. In fact, my plan, which has been approved by my supervisory board, is to expand our workforce by about 39 people through 2016 on qualitative areas, not related to this, in connectivity and other parts of the business. I am fighting very hard to make sure that that loss is not the case. However, the management control and cost base of our business in the UK is benchmarking the world, so I have some headroom to ensure that we are well supported and well resourced financially to make sure there is no direct effect in the UK. I must say, however, that during October and November I was concerned, but now that the orders have come back again I am really quite confident. We are still not through it yet. We have a lot of work to do with customers. I have to submit to that. We have to be clear, clear and clear again.
Iain Stewart
Have any members of staff wanted to leave the company as a direct result of the scandal breaking?
Paul Willis
Not to my knowledge. I have discussed it very informally with all levels of staff. No one has said to me that they want to leave the company. In fact, I would say it is the opposite. We feel that we are a little bit under pressure and that we have to work together to make our points. Of course we have to answer your questions, but everyone is working together to try to make the points as clear as possible and to be as open as possible to customers. That is our priority, and my staff in the different customer areas have done an absolutely outstanding job.
Martin Vickers
I am becoming more and more confused, Mr Willis, as we go through this.
Paul Willis
I am very sorry about that.
Martin Vickers
You have a device that you say actually does conform to the regulatory regime, yet you have written to 1.2 million customers. You have 450 staff working on it and you are analysing 40,000 responses. Why? You could not answer Mr Stringer’s question about how much it is costing you, but it is costing you an enormous amount in reputation. Would you agree that it has been a public relations disaster?
Paul Willis
It has certainly not been helpful to our brand image—
Chair
That must be the understatement of the year.
Paul Willis
But although consumers in general may not listen to me and they may not listen to the questions here, they read the independent magazines. If you read the independent magazines they make it very clear, and that is why customers are returning to buy our vehicles. However, we have a duty to the 1.2 million customers we sold the cars to, and we have to fix the cars so that they are within the regulations.
Chair
In your most recent letter to me, Mr Willis, you say that it is common across the auto industry for software to be installed that recognises when the vehicle is undergoing testing. What did you mean by that, and how do you know about it?
Oliver Schmidt
European regulation gives you the ability to recognise that your vehicle is undergoing testing in order to have repeatability of emission testing, so that every time you go through an emission test you achieve the same result and are not bound to random circumstances. That is in the European regulation; you are allowed to recognise the test. What you are not allowed to do is tamper with your emissions while recognising the test.
Chair
When you said that, were you reflecting on other places where this has been happening? That is not what the letter said to me. The letter did not draw attention to a regulation. It spoke about what it said was common practice elsewhere.
Oliver Schmidt
That is what I said. European regulation allows you to do it.
Chair
Yes, you have told me that it allows you to do it, but I am asking you where else it was being done, or is it being done?
Paul Willis
We don’t understand. What do you mean by “where”?
Oliver Schmidt
Do you mean at other manufacturers?
Chair
I am asking you, because you told me that it was common across the auto industry for software to be installed that recognises when the vehicle is undergoing testing. I am asking where else that was happening—it may still be happening now. Which other manufacturers were doing things like that?
Paul Willis
We do not know which other manufacturers, but we know it is common practice.
Chair
How do you know?
Paul Willis
Because the test is set up in a certain way. It is as simple as that. The test does not represent the real world.
Chair
What are you telling me—that the statement you made in the letter is not really correct, or that it doesn’t mean what we think it means?
Paul Willis
No, the statement is correct.
Chair
The statement does not refer to what is allowed. It talks about common practice.
Paul Willis
Would you mind just pointing the statement out to me so that I can refer to it and refresh my memory? Is it 21 December?
Chair
We will find it. It is the last letter you sent me.
Paul Willis
Which page?
Chair
You are telling me that in your view this is in the regulation, but you do not know if it is actually carried out by anyone other than Volkswagen.
Paul Willis
I don’t know who does what. I can only comment on VW.
Chair
So you do not know. Was the defeat device software that you admitted to using in the US market the same software that was installed in the European market?
Oliver Schmidt
No.
Chair
Why were there different types of software?
Oliver Schmidt
The base function of the software is the same—it recognises drive traces—but the effect of what the software does afterwards is different. In Europe, in an effective Euro 5 engine there is no NOx aftertreatment system on the engine. You have only an oxidation catalytic converter and a particulate trap for particulates. There is nothing that reduces NOx on a Euro 5 engine. On a US engine in our older generations you have a lean NOx trap. In the newer generations you have an SCR, a selective catalytic reduction system. In the US, the software influenced the function of those two systems, which are not present in the EU.
Chair
Don’t you think that somebody might think you were perhaps trying to deceive different regulatory authorities by acting differently in different places? Both of them were found to be wrong.
Oliver Schmidt
I do not understand the question.
Chair
Could that be interpreted as an attempt to deceive the different regulatory authorities, the European ones and the American ones, by making different adjustments in different places, and both have been found to be wrong, if not illegal?
Oliver Schmidt
I don’t know what to say.
Paul Willis
The question of deceit will be clearly brought out during the Jones Day inquiry. I am hoping to find out what exactly the reason was for this. I agree with what you are saying. We will find out in the Jones Day report.
Chair
If it was correct that a small group of engineers took a decision in Europe to take actions and install this defeat device, would you not then have needed another group, perhaps a different group of engineers, working on vehicles for the American market to take an equivalent decision there?
Oliver Schmidt
It is the same group.
Chair
It does not sound very likely, does it?
Oliver Schmidt
It is the same group.
Chair
The same group, if they were responsible, took a decision to do one thing to deceive authorities in one place and another thing to deceive authorities in another place. Is that what you are saying?
Oliver Schmidt
At the moment it looks that way, but, as Mr Willis pointed out, we have to wait for the details. The base software is developed in one department.
Robert Flello
Mr Willis, previously you told us that you thought most vehicles in the UK or in Europe would need to be amended with the software change, but there would be some that required a hardware change—a physical change.
Paul Willis
Yes.
Robert Flello
How many vehicles have so far been physically changed?
Paul Willis
None.
Robert Flello
Why is that?
Paul Willis
Because we have always said that the timeline starts in week nine, month three, and that is what we are working towards. I can report one small thing. Last week on the 2 litre engine in the Amarok, which is a very small volume, we eventually agreed the technical fix with the KBA. The engineers are very clear that there is no difference in the torque curve. There is no difference in the driveability or fuel consumption. In principle, the KBA agree and we are just waiting for the written signature this week. That is one piece of good news for everyone. The real volume you are talking about—the 2 litres—starts in week nine.
Robert Flello
Let me come back to the question. Have you identified how many vehicles will need a physical change rather than a hardware change?
Paul Willis
Yes. The physical change, which is the flow transformer, affects 430,000 vehicles, but that fix will not start until the ninth month.
Robert Flello
Mr Schmidt, as an engineer, tell me what the flow transformer does.
Oliver Schmidt
It sounds simple. In the air flow from the front end of the car to the actual intake manifold, you have an air mass sensor where you can measure the amount of air going in. In the configuration that was brought into production, the onboard computer could not get a proper reading of the air mass going in, so the actual volume of air mass going in was not what we read on the air mass sensor. The transformer, or rectifier as it is sometimes called, gives you not a turbulent but a laminar streaming around the sensors, so now we are able to get a proper reading of the air that is going in and we can inject the proper amount of fuel according to the air mass.
Robert Flello
What would happen to those vehicles if that piece of kit was not fitted—if just the software was changed to get rid of the defeat device but that piece of hardware engineering was not added?
Oliver Schmidt
That is a very good question. That is what we will be doing for the Volkswagen Caddy. That is certified under the N1 truck certification to be able to fix those vehicles earlier. We will put the fix—
Robert Flello
What would happen if you didn’t?
Oliver Schmidt
You would meet N1 regulation but not M1 regulation. The N1 regulation for the Caddy is 235 mg NOx and for the passenger car—the M1—it is 180, so it gives you at least an additional 55 mg per NOx reduction.
Robert Flello
So that I am clear, upgrading the software and removing the defeat device but not putting in that particular air flow device would increase the amount of NOx emissions, because there is extra fuel going into the engine that should not be, because you cannot get a proper reading. Is that correct?
Oliver Schmidt
You cannot say “increase”, because the actual test result was achieved using recognition of the drive trace. You cannot compare.
Robert Flello
Why can’t you compare? I am driving along in one of these vehicles now, with a defeat device fitted to the software, and it then gets plugged in and as far as I am aware it is emitting at the right level. If I now take it in and have the software upgraded to remove the defeat device, but do not do the hardware fix and plug it in, surely it is now going to fail the emissions.
Oliver Schmidt
That is correct, but it is not how the service will actually be carried out.
Robert Flello
I am sure it would not be. I am sure you would change the hardware as well, but in terms of the emissions, without that defeat software and without having to make a physical change to the engine, it now no longer complies with the emissions test.
Oliver Schmidt
It would not. Yes. As I mentioned, it complies with the N1 regulation for a truck, but it would not comply with an M1 if it is the same vehicle size. Of course, if you go to smaller and lighter vehicles, we have not measured those without, and I cannot give you exact results; this would be the comparison for a Volkswagen Caddy, which is one of the heaviest cars we sell—
Robert Flello
We are back to square one again, though, in that it is not just a simple matter of saying that without that software it would still pass the test anyway. It is not; you have to put a physical device in half a million vehicles.
Oliver Schmidt
On the 1.6 litre you are correct. We need to put in the flow rectifier in order to do that. There are even more changes. I compare it to an iPhone, for example. On the 2 litre and the 1.2 litre, we put new apps on your iPhone. On the 1.6 litre we give you a new operating system. The base software in the ECU is totally changed, while on the 1.2 and 2 litre, we are only changing the calibration data—although “only” is the wrong word. We do not change the base software, except of course for removing the switch logic—
Robert Flello
I hear all that, Mr Schmidt, but it still comes back to the point that for half a million vehicles, without a physical change to the engine—not just a software change—those vehicles would no longer meet the emissions test. That seems to fly in the face of what we have been hearing for most of today—that it is just a software problem.
Oliver Schmidt
As a theoretical game, yes; if you played that, you would be right.
Mary Glindon
The purpose of the emissions limits is to protect people’s health. Does Volkswagen engineer its vehicles for that particular purpose or is it only to comply with the letter of the law?