Transport Committee
Oral evidence: BMW vehicle recall, HC 1140
Tuesday 26 June 2018
Ordered by the House of Commons to be published on 26 June 2018.
Members present: Lilian Greenwood (Chair); Jack Brereton; Steve Double; Luke Pollard; Iain Stewart.
Questions 1-146
Witnesses
I: Graeme Grieve, Chief Executive, BMW (UK), Klaus Kompass, Vice President Vehicle Safety, BMW Group and Johann Ebenbichler, Vice President Quality, BMW Group.
II: Gareth Llewellyn, Chief Executive, DVSA.
Written evidence from witnesses:
- DVSA
Witnesses: Graeme Grieve, Klaus Kompass and Johann Ebenbichler.
Q1 Chair: Welcome and thank you very much for coming along today. I apologise for the slight delay in starting. For the record of our proceedings, could you introduce yourselves and the organisation you represent?
Klaus Kompass: Good afternoon. My name is Klaus Kompass. I am responsible for vehicle safety development for BMW in Munich, Germany.
Graeme Grieve: Good afternoon. I am Graeme Grieve, CEO for BMW UK Ltd.
Johann Ebenbichler: Good afternoon. I am Johann Ebenbichler, Vice President Quality, BMW Group.
Q2 Chair: We would like to begin with questions about the fault with the B+ connector. For the benefit of everyone here and those watching elsewhere, can you explain what the B+ connector is and what happens when it fails?
Graeme Grieve: If I may start with a simplified explanation, then I will invite Professor Kompass to give a more technical one. The B+ connector is a connection on the vehicle that joins the power from the battery, which is normally in the rear of the vehicle, to the fuse and control system in the front. It is the main power connection for the vehicle. Professor Kompass is an expert in this area and can give you a more detailed explanation, if that would help.
Klaus Kompass: The failure that occurred was related to vibrations we have during the lifetime of the vehicle, in combination with the amount of electric load that goes through the connector. With those two factors, plus time, the connector can produce an oxide layer that increases the resistance of the connector in that area and thus leads to a greater loss of electrical power.
Q3 Chair: BMW, as I understand it, first became aware of the issue with the B+ connector in 2010 but does not seem to have informed the DVSA until 2014. Why were you so slow to report that issue to the DVSA?
Graeme Grieve: When we were first aware of the issue, we believed that this was just a minor corrosion that resulted in some long-term deterioration of the connectivity. We didn’t realise at that stage that the issue would become so significant and therefore we did not inform the DVSA, as we did not believe at that time it was a safety-critical item. The first time the DVSA was aware of the issue was when we informed them of a recall for cars with what we term a “hot country specification” in March 2013. The majority of those vehicles were in north America, in Canada. Broadly that specification means that there is a more powerful fan in the car, with a higher electrical load, and that showed the problem first. There were 78 vehicles in the UK with the same specification and they were recalled at the same time. That was the first time the DVSA was informed about the topic.
The first time the DVSA received a vehicle safety defect report was in October 2014, which we investigated together with them.
Q4 Chair: Two questions arise from that. First, given the description you gave us of the fact that this provides the electrical power, why didn’t you realise at that point that a fault in it could lead to the car losing all power, including its lights, windscreen wipers and so on? Why was that not clear from the initial fault?
Graeme Grieve: From a consumer perspective first of all—I am sure Professor Kompass will add something—we believed from our information that the degree of corrosion and the low speed of corrosion would mean the driver of the vehicle would experience a certain number of warning signs. Therefore we would introduce what we term a PuMA measure, which is a technical action system that is available to our retailers and to independents. That would allow them to diagnose some of what we term signs of prior warning or alert, like the central locking failing to work or the car failing to start. At that stage, the B+ connector could be examined and, if the corrosion was sufficient, replaced. We did not believe and understand at that stage that the corrosion could result, in the long term, in a loss of power. We believed that where customers experienced those signs of prior warning they would take action. The majority of drivers did, but clearly some did not and we did not foresee that at the time.
Q5 Chair: I think we will come back to that issue in a moment. The DVSA thinks that you were aware of the B+ connector issue as early as 2011. Your evidence to us said that you were first aware of the problem in July 2010. I appreciate that you have said that you did not appreciate the significance of the fault, but did you give them and the coroner the wrong date? How do you explain that discrepancy?
Graeme Grieve: No, I do not believe that we gave the wrong date. The first case we were notified of by the DVSA was in March 2013, which we evaluated together with them. We provided details of our PuMA analysis, and we provided details of the past sales to date, which were requested by the DVSA to give them an idea of scale. What we did not do at that stage, which we do now as part of the learnings and the change of process, is we did not interrogate our warranty system to see if there had been any previous reports of a vehicle with the same fault reporting a loss of power.
There are a number of learnings from this incident. The other, which we may discuss is blower regulators, which have now been corrected and implemented within our organisation. At that time, we did not withhold any information from the DVSA; we were simply not aware of the first instance in the UK until the DVSA’s safety defect report in March 2013.
Q6 Chair: But were you aware of there being problems as early as July 2010?
Graeme Grieve: We were aware of the potential of corrosion resulting in a loss of connectivity in July 2010, which is why a PuMA measure was implemented—again, the service diagnostic system—to inform our retailers that if a customer presented themselves with any of these issues, such as failure of the central locking or failure to start, then one of the items they should check on the car was the B+ connection.
Q7 Chair: Okay. What did you do once you received notification of the accident in December 2016 caused by a faulty BMW in which Mr Gurung lost his life and his wife was seriously injured?
Graeme Grieve: We had a meeting with the DVSA in February 2017, which was the first time the DVSA informed us that they believed this to be a safety-critical issue. There had been ongoing discussions with the DVSA since October 2014. At those discussions in February 2017, we still believed that the vehicle defect was not safety critical, but in discussions with the DVSA we changed our view and agreed to a recall of all petrol-engine vehicles at that time. That recall paperwork was submitted and the recall was live from April of that year.
Q8 Chair: So in February 2017, two months after Mr Gurung lost his life, you still did not think it was a safety-critical issue, yet the circumstances of Mr Gurung’s accident had been foreseen in a meeting with the DVSA in February of that year—of 2016—where I understand one of your engineers made a note saying, “We don’t want a fatality.” Surely you should have realised that this could have the potential for someone to be killed or seriously injured at that point.
Graeme Grieve: At that point we still believed that sufficient prior warning would be provided to the driver. Also, in terms of our RAPEX assessment, which the Committee may be familiar with—a European standard risk evaluation that is used in the industry, which my colleagues can explain—this had a low risk. However, in discussion with the DVSA—again, this was the first time their opinion was shown to us that there should be a recall—we agreed to a recall, and that was launched in April of that year for petrol-engine vehicles.
Q9 Chair: Couldn’t action have been taken more quickly, either in terms of accelerating work on the recall or issuing advice to owners?
Graeme Grieve: We believe that from the date of the recall being communicated we did everything we could to ensure that on a speedy basis vehicles were attended to and repaired.
Q10 Chair: Let us look at some of the international issues. As I understand it, BMW recalled cars for power loss in the US, Australia and South Africa in 2016. When was the first recall initiated for the B+ connector issue anywhere in the world?
Johann Ebenbichler: (Translation) The first recall for B+ was in January 2013. Prior to that, we discussed the case with Transport Canada, which is the trading standards agency in Canada, and NHTSA, the trading standards agency in the USA. We had an internal review, and then we had a global recall for 800,000 vehicles. We were of the opinion that the issue had been dealt with, but over time we realised that more cases kept occurring in other countries, so we had the second technical action in October 2016. 172,000 vehicles were affected, out of which 36,000 were from the UK. The first recall was in the hot climate countries, as my colleague described. Out of that, 78 vehicles were from the UK. Whenever it is necessary and we see it is an international problem, we do global recalls, not just national recalls.
Q11 Chair: So you realised that there was a problem in the hot climate countries earlier than 2016—in 2013. What was the reason for the delay in the recall in the UK?
Graeme Grieve: Perhaps I can invite Professor Kompass to explain the difference in the electrical architecture of the vehicles. That may provide some additional background.
Klaus Kompass: The hot climate package contains a much stronger ventilator, and the power load is twice as high as for cars without the hot climate package. As I said at the beginning, it is vibration plus the amount of electrical power that is being transferred through the connector, so the higher the electric power that goes through, the higher the probability of these flaws.
Q12 Chair: When did BMW make reports up to the other competent authorities in EU member states and to regulators in other countries?
Johann Ebenbichler: (Translation) In October 2012, we had a discussion with NHTSA, the US agency, and Transport Canada. We started discussing how risky this issue was. At the time, with the high fault rate and complaint rate, we decided to go for a safety recall. It is first always important to liaise with the agencies and authorities to get their opinion and share our opinion, and then come to a decision about whether it should be a safety-relevant recall or a service campaign that goes through the dealers. I assure you that if an authority tells us it is safety relevant, we will go for a safety-relevant recall, because our customers’ safety is essential for us.
Q13 Chair: So that was October 2012—a long time before the UK recall, which was not until February 2017. In that period of five years, was there nothing that could lead you to believe that this was not just a problem in hotter climates, but that there was a wider problem across BMW’s full range of vehicles?
Johann Ebenbichler: (Translation) We have sold 20 million vehicles worldwide, and we have 4,400 warranty claim dealers. We look at the cases globally, of course, but we rely on good reporting from customers and dealers. As Mr Grieve said, the first case in the UK was—when was the first case in the UK?
Graeme Grieve: October 2014.
Q14 Chair: That is still quite a long time. In October 2012, you realised there was a problem in the US and Canada—well, that’s when you had a discussion with their authorities. In 2014, you realised that this is also a problem in the UK, and yet it is a full three years before there is a recall. Is that right?
Graeme Grieve: That is correct, Chair. If you would allow me to give a short chronology of the period between those two events, that may assist. October 2014 was the first vehicle safety defect report that we received from the DVSA. As you would expect, we investigated it together with them. The information we provided to the DVSA at that stage seemed to satisfy them that this was non-safety critical. The next time the discussion with DVSA took place was in February 2016, where again we reviewed the cases. We therefore presented at that meeting 20 additional cases, which we had found from our research had previous warranty claims that had indicated a loss of power.
At that stage, both BMW and the DVSA still did not consider that to be a safety-critical issue, but we agreed with the DVSA to monitor on an ongoing basis, which we did until June 2016, when the decision was taken with the DVSA to close that monitoring period. Those events happened before the first event in 2014 and the recall in 2017.
I understand that it may seem strange, but at that stage both BMW and the DVSA did not believe during that period that this was a safety-critical item. The only occasion when it became clear that the DVSA had changed their opinion was in early 2017, after the tragic death of Mr Gurung.
Q15 Chair: But I do not understand how it can be possible that one of your engineers can sit in a meeting with the DVSA in February 2016 and make a handwritten note that makes a reference to the potential for a fatality and for you to tell me it took you a whole year to realise that it was a potential safety issue. Your engineer who sat in that meeting in 2016 recognised that there was a safety issue, didn’t he?
Graeme Grieve: Chair, the note in our engineer’s diary, according to our engineer, was a statement that was made by a representative of the DVSA; it was not his own. The only notes we have of the meeting that relates to that point are our own in the engineer’s notes. That is something I have to accept the opinion of our engineer on.
I am not denying that the statement was made, but I believe it was made by a DVSA representative, not by our engineer. Because, in early 2017, as I mentioned, when the DVSA first approached us to say that they believed that this was a safety-related item, at that time, based on our RAPEX and belief that the signs of prior warning would be heeded by the customer, we did not initially agree. We changed that opinion, clearly, and we submitted our recall paperwork in March of that year, and the recall was launched in April 2017.
Q16 Chair: I want to come back to the issue of prior warning a little later. What specific actions do you take to encourage good reporting from customers and dealers across the globe and ensure that you are picking up problems?
Graeme Grieve: If I answer first on behalf of the UK, I am sure my colleagues will pick up. We have a number of feeds of information. We have a feed of information that comes from our approved retailers via a system we call PuMA, which records all technical actions and complaints.
We also monitor customer complaints that come directly into our organisation. We now see some changes we made in 2017 and we also monitor social media to see if we can understand if there are any problems we are not aware of, or current trends. PuMA is a global system that our colleagues in Munich have visibility to. That would allow them to summarise and analyse complaints and these issues on a global basis.
Q17 Chair: Have you taken specific actions to improve that level of reporting as a result of some of these issues?
Graeme Grieve: Yes, we have, Chair. There have been many learnings in this for us as an organisation and we have made a number of changes. If I may summarise some of them, first, in terms of dealing with fire, which is not related to B+, but was clearly a feature of the Select Committee investigation of Vauxhall Zafira, before that investigation we had followed the general industry practice of referring any case of fire to the insurer. If the insurer wished us to be involved, we would. We changed that from August 2017. Now, not only would we investigate any fire, but we would actively pursue inspection of that vehicle either jointly or independently from the insurer, so that we understand as early as possible if there are any technical issues that contributed.
Sitting suspended for a Division in the House.
On resuming—
Q18 Chair: Just before we broke for the Division, I think you were explaining the changes that you had made specifically to encourage good reporting from customers and dealers. I do not know whether you had finished telling me about the specific actions. Let’s go back to that.
Graeme Grieve: No, I had not finished. We segued from the reporting of data. I talked a little about the PuMA system, which is a global solution. Colleagues can add more on that, if it helps. As far as the UK is concerned, we have taken many learnings out of the B+ connector discussions. We have made a number of fairly important changes in the UK already.
We have some very important feeds of data and information. Our retailers who do the normal repairs use the PuMA system to feed back to us technical measures and instances and problems that they have found. We also collect information from our customer service team. If a customer complains directly to us, or via one of our retailers, about a problem, we collect and analyse that information. We also established a team from August 2017 to manage all those feeds of information where there could be a safety-critical issue and also to effectively scour social media. We did that because we established that sometimes the problems were being discussed on forums and social media before we were even aware. If a vehicle visits an independent retailer, we may not be aware of the issue that the driver or the customer has encountered.
One of the very specific changes as a result of that was with how we manage cases of fire. Prior to the Committee’s review with Vauxhall, we would have followed what I believe is industry practice, which is we would have simply referred customers who had experienced fire to their insurers. We would have got involved willingly if the insurer had asked us to because they believed a vehicle defect was present. Since August 2017, we now actively seek an inspection, either jointly or independently with the insurer.
One issue I would like the Committee to consider is having more transparency in future between the insurance industry, manufacturers and the DVSA. There could still be instances of fire that we are not aware of as a manufacturer and that the DVSA is not aware of. There is no set process for insurers to inform either the DVSA or us of cases of fire. That would be an enhancement.
Another change we made is that all the ongoing safety matters we are dealing with and are investigating are reviewed regularly at our UK management committee, which happens every two weeks. The changes we have made certainly make our processes more relevant and robust for the future.
In terms of the broader global information on technical reporting, we do that reporting. This is an area that Johann is an expert in, if it helps to give more background on how we use that data on a global basis.
Q19 Chair: I think what would be helpful to know is whether, as a result of the specific changes you have just described, you have seen an improvement or a change in the levels of reporting.
Graeme Grieve: As far as the UK is concerned, yes, we have. It has allowed us to identify some issues much earlier that we might otherwise not have been aware of until later. None of those are yet safety critical. We have some areas under investigation and evaluation at the moment. But we certainly have, effectively, the taskforce approach. Previously, we had the technical team based in the UK. They would handle a myriad of these technical inputs by PuMA, whether it was a simple case of a light bulb not working in the interior or a potential safety-critical issue. Now those are focused into this team to give them the attention they deserve. It also greatly helps us in terms of how we then present and discuss those issues with the relevant technical team at the DVSA.
Q20 Chair: You mentioned independents, which I presume means independent garages that do some of the repair work. Do you have ways of working with them to ensure that they report things, even if they are not your own dealerships?
Graeme Grieve: That facility is available. We have a system called aftersales online that is available to any independent retailer, which gives them almost the same functionality that one of our approved retailers would have. That is free to join, but there is a subscription for usage that ranges between £250 for someone who is using it occasionally and £3,000 for someone using it regularly. The key issue for the industry, though, is that that is what we do as one manufacturer. Many other manufacturers provide the same services, but, if you are an independent, quite often you will be servicing a range of manufacturers’ products and marques. It is unlikely, as an independent doing that, that you would access a multiplicity of manufacturers’ systems. There is a role here, perhaps for the Committee or the DVSA, to look at a UK system that would allow us to combine that information.
At the moment, independents can look up outstanding recalls via the DVSA system, but they have no access to other technical issues that are under evaluation. To be clear, these would not be safety critical, otherwise they would be subject to recall, but, in terms of the diagnostic process, that perhaps could help the independents.
Q21 Iain Stewart: Can you tell me how many reports of problems with the B+ connector you have received?
Graeme Grieve: So far, we have received 106 complaints relating to the B+ connector where the customer has experienced stalling or a cutting-out of the vehicle. That would be the critical number.
Q22 Iain Stewart: But the total number relating to this?
Graeme Grieve: The total number for B+ is very hard to give. I will explain why. Because the PuMA service action was in place, a lot of the customers experiencing a problem with the B+ connector would have visited their retailer, and it would have been replaced as a normal course of action. So we have part sales data, and the data that we have up until the first technical campaign, which was late 2016, showed that we had sold 3,200 of those cables, which would either have been as a direct replacement or might have been the result of accident damage or some other area. What we do not know specifically is how many of those cables were sold because the customer had experienced the connectivity issue. What we do know is that 106 drivers up until the end of May had reported either a stalling or an interim or permanent loss of power.
Q23 Iain Stewart: Of those 106, how many occurred without any warning?
Graeme Grieve: We believe that there would always have been a prior warning. One of the key learnings of this for us is that in many cases customers did understand the prior warning and have the vehicle attended to. In others they did not, so one of the learnings for us, perhaps even for the DVSA and the Committee, is how prior warning is interpreted by the manufacturers and ourselves going forward.
Q24 Iain Stewart: How many occurred when the car was actually in motion?
Graeme Grieve: The 106 are where the car is in motion, which would either have been on the road or could have been in a car park or any other situation. This is where the engine of the vehicle would be running.
Q25 Iain Stewart: In May 2016, you told the DVSA that the number of reports had fallen significantly, but that is not supported by the coroner’s findings. Do you have any evidence you can provide to support your claim that that level was falling?
Graeme Grieve: I can easily write in with whatever evidence the Committee requires. As part of the review with the DVSA, which was cleared in June 2016, we had received an additional five vehicle safety defect reports.
Q26 Iain Stewart: Turning to the issue of the blower regulator, how long was it between your first becoming aware of the issue and informing the DVSA about the potential defect?
Graeme Grieve: The first time we were aware of the potential issue was when a PuMA measure was issued. Again, this is the global tech reporting system, which would highlight to us and the retailer if a driver experience was a problem, and it would give them some diagnostic routes. That PuMA was launched in November 2012. The first cases we were aware of in the UK, where there was seen to be some overheating damage to the wiring—so this is certainly not a fire, but discolouration of the wiring—were in January 2016. At that time we highlighted the issue to our colleagues in BMW AG in Munich—that we had seen this discolouration in wiring, that there could be some risk if it continued, and that it could lead to fire.
The first case of fire we saw in the UK was in January 2017. That was a report to our customer service team. As was our process at the time, we referred the customer to their insurer, who inspected the car. There was an independent forensic inspection of the car, which concluded it was likely to have been caused by a fault within the blower circuit, which could well be a problem we are talking about later. We did not inspect the car at the time, because at this stage our process was to ask the customer to discuss with the insurer, and we would only get involved if the insurer asked, which they did not on that occasion.
We had a second report of fire in February 2017. Again, the customer at that stage in the old process was referred to their insurer, and again we did not have the opportunity to inspect that vehicle. The first time that we had an active discussion with the DVSA was in April 2017, in response to a vehicle defect report—again, with the DVSA concluding in discussion with us that the root cause at that stage was unknown. So April 2017 was the first time we were in active discussion with the DVSA regarding the heater-blower regulator.
Q27 Iain Stewart: But April ’17 was, by my calculation, 15 months after you were first aware of the overheating problem. Was there never a discussion within the company, given the other issues, that this ought to be reported at an earlier stage?
Graeme Grieve: No, because when we saw the overheating in the wiring, we reported it to our experts in BMW AG, because we had seen the overheating. Again, it is only three cases of overheating—a relatively small sample. We thought if that continued, there could be a risk of fire. At that stage we did not feel that that was necessary to report to the DVSA, because of the fact that we had seen that discolouration. We had not experienced any cases of fire at that stage.
Q28 Iain Stewart: Perhaps one of your colleagues could explain why there wasn’t a decision made at the parent company level to inform either the DVSA in this country or equivalent bodies elsewhere in the EU.
Johann Ebenbichler: (Translation) In 2011 we saw very few cases, and the few cases of the blower regulator were in countries where it was running at top level. That was the problem. Mr Kompass can explain this as well. The problem occurs if the blower is used at the highest level for long times. We primarily saw it in Egypt, Taiwan and the US. All in all, we had 150 cases out of 5 million vehicles. Those 150 cases out of 5 million globally were mainly in the hot countries I have just mentioned, and a very small share were in the United Kingdom.
Q29 Iain Stewart: But given the other problems that you were noticing, did you not think that you should err on the side of caution and notify the DVSA and equivalent bodies earlier than you did?
Johann Ebenbichler: (Translation) In the most affected countries, we did get in touch with the authorities and discuss it—in the countries where we had seen the most cases. In the UK, once we had a case, as Mr Grieve just explained, we got in touch with the DVSA. In October 2017 we decided that we would have a recall affecting 1 million vehicles, because we thought it would be wider than we had thought. That was then a global recall.
Q30 Iain Stewart: Just for clarification, you notified the DVSA in April 2017. When did you notify some of the other countries’ regulatory bodies?
Johann Ebenbichler: (Translation) We talked with the US authority NHTSA in autumn 2017. At the time, there were fewer than 100 cases globally, with 5 million vehicles.
Q31 Iain Stewart: So that was after the UK regulator was informed. You are saying the autumn of 2017, when the UK was informed in April 2017.
Johann Ebenbichler: Yes.
Q32 Iain Stewart: So no decision was made at the parent company level, looking at the data from the UK and elsewhere, to notify any regulator earlier than the UK.
Graeme Grieve: Just to be clear for the record, in April 2017 we were in a discussion with the DVSA relating to a vehicle safety defect report that the DVSA had raised. We did not inform the DVSA at that time that we believed that this was a more significant issue, but we had an active discussion with them regarding the vehicle safety defect report.
Q33 Iain Stewart: So it was later in 2017 that you notified the DVSA that this was indeed a serious issue.
Graeme Grieve: Yes. Following the recall for North American cars, as we always would, we informed the DVSA of a recall elsewhere in the world so they were aware, and we informed them at that stage that it was a current investigation in the UK.
Professor Kompass can add something. Again, there is an element around the technical architecture here. We believed at that stage that the fault was more likely to occur in the US. Again, it is subject to electrical loading, particularly as the fan motor is on maximum speed, which is more likely to occur in the USA. For some technical reasons we believed that the issue was more likely to occur in the USA, but we did, from that point, have an active investigation in the UK, and we then concluded in May 2018 that a recall was required on 3 series vehicles in the UK equipped with that component.
Q34 Jack Brereton: I want to follow on from Iain’s questions, particularly with regard to the B+ connector recall. Alex Neill from Which? has suggested that it took several more years than in other countries to inform the UK regulator and to commence the recall in the UK. Why was that?
Graeme Grieve: As I mentioned in earlier evidence, we have been in active and regular discussion with the DVSA since the first vehicle safety defect report in 2014. Up until the point of the discussions with the DVSA in early 2017 we did not believe that this was a safety-related defect. We believed that there would be sufficient signs of prior warning via the central locking not working or the car not starting that the driver would seek assistance and have the car inspected and repaired. It is clear with hindsight that although that happened in the majority of cases, some customers did not understand those signs of prior warning. As I mentioned earlier, one of the other learnings in this for us—
Q35 Jack Brereton: So you are blaming it on those customers, then?
Graeme Grieve: No, I am not blaming it on the customers at all. If you would allow me to finish, I think one of the learnings is the interpretation of prior warning. We had very much taken the view as experts and engineers that those signs of prior warning would be a clear indication to the customer that they need to get that car checked, inspected and repaired. A number of customers did not heed those warnings. I believe we need to take a far more consumer-centric and driver-centric—
Q36 Jack Brereton: Why did you think it was sufficient to order a recall in some countries, but it took several years for a recall to commence in the UK?
Graeme Grieve: Regarding the B+ connector, when was the first recall, Johann?
Johann Ebenbichler: (Translation) The first recall was in October 2012, in the context of many cases involving the SA 823, which is an American model. So the problem first presented itself on a larger scale in the USA; in the UK it was not visible to us at that time. That is why we recalled 800,000 vehicles with the specific model SA 823 globally, and then only later did the problem increase in the United Kingdom. Then you had a second recall in October 2016, when 170,000 vehicles were recalled globally and we saw that the problem was increasing. That’s why we reacted.
Klaus Kompass: As I said before, the amounts of electrical power and electrical current are very important aspects in producing and accelerating the corrosion. All US cars have the hot climate package with the bigger fan—even though there are some areas in the United States that are, from a temperature point of view, the same as the UK or Germany—so they consume more electrical energy. They consume twice as much electrical energy, and twice as much electrical power goes through this connector. That was the first indication that this was the major root cause.
It finally turned out that over time—and this is probably something that we underestimated, I have to admit—cars with a lower amount of current could also see that effect. It was very few of those cars, but still, some of them could also experience that. That was a learning that we took afterwards, and we acted on it.
Another important point is that we talked about the prior warning aspects, and I think it is important to know why we saw so many situations where a standing car shows more effects than a driving car. That is based on the fact that the temperature of the connector goes down as the car is standing still. The engine is off, and very little power is going through that connector, so the connector is cooling down. As the connector cools down, the pressure of the pins pushing against each other has been reduced, which increases the resistance and increases the likelihood of those instances occurring.
That was frequently the situation with those rare events and rare incidents. The majority of those incidents were situations where people approached their car, were trying to open it, and saw that the central locking didn’t work. They were trying to start the engine and the engine didn’t want to start. Much less frequently, while the car was driving, there was some power cut or some engine cut for a short time, and most of those cases were non-critical.
We always called those “prior warnings”. I prefer to call them “prior incidents”, because we didn’t expect consumers to understand that if you cannot open your car, this might produce a safety-critical aspect at some time. Frequently, we experienced that, if this happened, people were simply annoyed. They went to the dealer: they wanted to get their cars fixed. That was a good indication for us that those prior warnings were sufficient, and they were really very helpful, because they brought the customers to the dealers to get the car fixed.
Q37 Jack Brereton: Johann suggested that the American recalls in 2012 were recalls of a slightly earlier model, and that there were potentially differences in the components. Could you explain what the difference is—if there is a difference—between the components used in those models in the US and the models in the UK that we had issues with?
Klaus Kompass: It was the cars with the hot climate package.
Q38 Jack Brereton: So there isn’t any difference in the components.
Klaus Kompass: No. The components, in general, are the same, but the amount of current is not the same. We had 78 cars in the UK that were also recalled in this first US recall.
Q39 Jack Brereton: So why was it not envisaged that we would have the same problem in the UK as we had with the models that had to be recalled in 2012?
Klaus Kompass: Because of the lower power load, except for the 78 cars in the UK that were recalled. Those cars also had the hot climate package.
Chair: Thanks. Luke, we will pass over to you to ask some more questions about the decision making that the company adopted.
Q40 Luke Pollard: I drive a Mini, which is part of the BMW Group. For the benefit of those people watching who might not be as familiar with all the details, when you talk about the models affected by the B+ connector, what model cars are you talking about?
Graeme Grieve: The B+ connector does not affect Mini. It affects BMW 1 series vehicles, 3 series vehicles, Z4 and X1 with a production range between March 2007 and September 2011.
Q41 Luke Pollard: I think that will be useful for people trying to work out whether it affects their car, because some of the experiences you have spoken about are quite familiar to many of us who drive cars—having a warning light pop on and wondering what it is about.
Graeme Grieve: What we have done as part of our communication process is that on the front page of our website—admittedly on the BMW website, not the Mini website, because we did not think it was relevant—you can access details on the recalls, which allow you to determine whether your vehicle was affected or not, by chassis number.
Q42 Luke Pollard: I am keen to ask you some questions to understand your decision making and when and why you took the decisions you did. Why did BMW initially think a non-code action was the right way forward with the B+ connector issue?
Graeme Grieve: We believed at the time that the issue was non-safety critical, otherwise we would have pursued a recall with the DVSA. We did, however, believe that the issue needed attention, and we followed an internal process called a technical action, so our retailers were aware and retailers would therefore be able to contact any customers that they had on their data file. That clearly does not allow us to reach all owners, because taking a car of the age at the time, they would not all be known to our retailers. Again, that is a learning, and in future we are more likely to approach the DVSA to conduct a non-coded action in that respect, to ensure that we reach all the customers affected.
Q43 Luke Pollard: When you issued your recall in 2017, what changed in your mind from dealing with it as a non-code action to dealing with it as a recall?
Graeme Grieve: Very much the discussion with the DVSA, where their opinion at that time was that they believed it was safety critical and requested us to do a recall. As I mentioned earlier, while we did not share the opinion at the time, we agreed with the DVSA and issued a recall notice in March of that year, effective in April. One area that I know has been discussed before is the voluntary code between manufacturers and the DVSA. I believe we have a very open and professional relationship. It is sometimes challenging, but we appreciate the expertise and knowledge on the DVSA, and if they feel so strongly about these issues, we would normally have a discussion, not only with the local technical team, but involving colleagues from BMW AG in Munich, to ensure that matters are fully understood. On this occasion, we agreed with the DVSA and therefore issued a recall notification.
Q44 Luke Pollard: At the point where you say you agreed that there is a safety-critical issue, was that because you had been told by the DVSA, “This is a safety-critical issue,” or had you come to that decision yourself at that point?
Graeme Grieve: No, we were led to that decision by the DVSA. At that stage, if we look at our analytical procedures and RAPEX, as far as we were aware from our internal decisions that would have still been a low-risk item; but the DVSA had a strong opinion that it was a recall and, following the tragic accident that resulted in the loss of Mr Gurung, we felt that, after discussion, it was the right thing to do.
Q45 Luke Pollard: After that decision, is there a point where you accepted that the decision was correct and had evidence yourself that that was the right decision and it was safety critical?
Graeme Grieve: The answer is yes. On the basis that we have now recalled a much wider range of cars, including diesel vehicles, it is clear that that was the correct decision at the time.
Q46 Luke Pollard: When you started your recall in February 2017, it only affected a sub-category of the vehicles that had been affected. Why did you only target a smaller group at that point?
Graeme Grieve: The recall was requested in March 2017, and we wrote to customers in April 2017. It always takes some weeks to get the data from the DVLA. At that stage it was related to petrol vehicles. The electrical architecture on the electrical and diesel engine vehicles is different; Professor Kompass can explain that in great detail. At that stage, we believed the issue was contained to petrol engine vehicles. We were asked by the DVSA, in April 2017 when the recall was live, whether we intended to recall the diesel vehicles. We said at the time that we did not, because of the difference in electrical architecture, but those were part of the broader recall in May 2018.
Q47 Luke Pollard: Then you extended it from petrol vehicles to other categories of vehicles?
Graeme Grieve: To diesel.
Q48 Luke Pollard: How many times did you extend the categorisation of vehicles included in the recall?
Graeme Grieve: There was effectively one extension, to use your term, but they were two separate recalls. The first recall related to petrol engine vehicles only; the second recall related to diesel and an extended range of production dates on petrol vehicles.
Q49 Luke Pollard: At the inquest, BMW conceded that there were reports of B+ connector failures without any warning. Do you accept that those reports undermine the argument that BMW use to justify dealing with this as a technical action?
Graeme Grieve: Yes, one of the learnings from this, as I mentioned earlier, is that we had relied on prior warning. We had relied on those alerts that Klaus mentioned being recognised by a customer as a need to have the vehicle attended to.
It is clear now, on the information that we have, that not all customers recognise that. That is certainly an issue we need to consider carefully, together with the industry and the DVSA, for the future. We would welcome a clearer definition of what would constitute prior warning, in relation to analysing a safety defect. But, at the time, not only was prior warning a consideration for us but also on the RAPEX evaluation, this was considered to be a low risk.
Q50 Luke Pollard: So when you get one of these concerns that we have been speaking about today, do you take these vehicles and test them yourselves? Do you try to test them to destruction and understand what the B+ connector problem or fault was?
Graeme Grieve: I will hand that question to Johann.
Johann Ebenbichler: (Translation) We have 4,500 dealers globally and they report problems to the centre. We look at the faults and defects and, if there is fault like the B+, we will go out in the field to see where the fault occurs, visit customers to see the vehicles and we also have vehicles on test benches centrally, to understand the fault. We had vehicles at that time and we had test-drive analysis.
As Mr Kompass said, you have a lot of connections flickering on the dashboard. That is something that customers might see and understand to be a prior warning. In 2017 we had a goodwill campaign. Goodwill means if the warranty has expired—you normally have two, three, four years 100% warranty for any cost that might occur—for those customers where that had already expired, we said they could still come to a BMW dealer to see what had led to the fault. So we tried to adhere to our duty of care for customers to see what the technical cause was.
Q51 Luke Pollard: Thank you. In crude terms, is the B+ connector doing what it was supposed to do? Is it a fault with the connector production, design, manufacturing? Is it a problem with the bits it connects into? Or was it doing exactly what it was supposed to and it was just a behavioural change in terms of the oxidisation that has caused these issues?
Klaus Kompass: In general, and in the vast majority of cases, the connector is doing what it is supposed to do. When we identified the first indications of customer complaints—for example, in those cases in uncritical situations where the customers could not open their cars—we immediately started a robustness measurement and changed the coating of the connector from tin to silver. That was in 2011; that was the first thing. That is the normal process. If we identify that there is some—
Q52 Luke Pollard: That is before the first problems were identified.
Klaus Kompass: That was when the first complaints occurred in uncritical situations. We did not identify those as safety critical, but we immediately changed the production process to a new coating of the pins.
Q53 Luke Pollard: You also extended the recall in 2018, didn’t you? You extended the recall to other vehicles.
Graeme Grieve: With a B+ connector, yes. That was effectively the second recall.
Q54 Luke Pollard: The second recall. We are trying to understand why it took so long. It seems from the timeline you have just been talking through that there were a variety of times throughout the six-year period, since it was first identified, when it was clear that there were problems with this.
Is it normal practice to shuffle up that little problem, whatever it may be, whether it is a fault with a hub, steering wheel or B+ connector? Is it just to deal with that one? Or is there a point at which you test each one and look at the implications of the problem? Were questions asked about what this problem will lead to? It seems as if you have done only a little bit—you thought it was a little problem so you took a little step forward and recalled as few cars as you could.
Graeme Grieve: I can you give you a UK perspective in answer to your question, then my colleagues may wish to come in. If we talk about the B+ connectors, which is probably quite an easy topic to cover, the issue is electrolytic corrosion of the contact, which means that connectivity can be partly or fully lost. That happens with time and is dependent on the electrical loading and the environmental situation.
When the first issue was determined, we thought that it was a quality enhancement issue, and the contact was therefore changed from tin to silver, which limited the electrolytic corrosion that would take place. As things progressed it became clear that the issue was continuing, and in a wider range of vehicles with a wider range of electrical loading. Again, the reason that this seems progressive—I understand how it may appear confusing—is that as the item progressed, both through the age of the cars and the degree of corrosion, we realised that, in vehicles with an architecture that we did not think would cause or trigger this problem to the same extent, it did, so we reacted in a progressive way.
That development is not unusual. It is quite normal for these problems to become more evident with time—not always, because some are clearly evident at the beginning, and we therefore take appropriate action then. However, in the case of the B+ connector, the significance of the impact of that corrosion and the degree of progression on that became clearer with time. That is why we have a stepped approach. We have now covered all the relevant vehicles that were affected by the B+ connector, so when the recalls are completed that issue should be finished.
Q55 Luke Pollard: At first you were looking at changing the material from tin to silver, as you just explained. This B+ connector is pretty critical in making the car work, by the sounds of it. Did no one say to hold on because, if it doesn’t work, the car will shut down or there will be a risk of it losing electrical power?
Does anyone within the organisation have the remit to challenge and to say, “Wait a minute; this could be a bigger problem than just the corrosion that we are dealing with”, and to think through the implications of that? I am just trying to understand the set-up of your organisation and whether there is that challenge within the organisation to think a couple of steps beyond the problem that you are encountering.
Klaus Kompass: If we identify a situation or incident like that, we normally start with a risk assessment right away. It is always very easy but very wrong to let your gut feeling decide whether it is safety critical or not. As Graeme already mentioned, we use the so-called RAPEX system. RAPEX might be well known to most of you as a rapid exchange of information system, but it also contains a risk assessment methodology.
That is the best methodology that we could find, and we have used it for many years now to identify the amount of risk that comes with those incidents. We take the number of complaints and the number of cases that we have identified. We then identify the root cause and look into the likelihood of that root cause happening in a critical situation, what kind of circumstances come with those critical situations, how those situations can be controlled and what kind of maximum outcome there can be. For example, a fatality increases the risk.
We use RAPEX but there are other methods. In the United States, the National Highway Traffic Safety Administration uses a slightly different method, and there are others in Australia and in China. We are very open to discuss that with everyone—with the Driver and Vehicle Standards Agency or anybody else who wants us to change that methodology. However, I am very sure that we need a joint risk assessment to identify the risks that come with each and every incident that we identify. Based on that risk assessment, the risk was calculated to be low in those cases.
Q56 Luke Pollard: Professor Kompass, you mentioned customers understanding prior warnings, and that a warning light is a sign of a bigger issue. The other day, one of my volunteers in Plymouth had a new warning light flash up on their Mini, and they said, “What does that mean?”, to which my answer is, “I don’t know.” I don’t know all the ins and outs, but like most motorists who see a warning light come on, I think, “There is a problem and I should probably take it in to get it dealt with.” Do you think that BMW has placed too much emphasis on the driver of the vehicle understanding what the flickering of warning lights means when in this case it is part of a much bigger electrical problem?
Klaus Kompass: The answer is twofold. First, if a red warning light is the only chance we have to communicate with the user of that product, the user should think, “Stop it right away.” In this case—this is a very important point that I mentioned earlier—I think we had an even more successful method: there was not a prior warning, but most customers reacted the way we thought and found it so annoying that they went to the dealer to get it fixed. The consumers did not necessarily have to think, “I see something that could be a critical situation,” because it was so annoying that they went to the dealer to get it fixed. They did not want, the next time they had an important appointment, to be standing in front of their car—
Q57 Luke Pollard: Effectively, you have relied on the annoying behaviour of the vehicle, which should not have been doing that.
Klaus Kompass: We did not rely on it; we saw that this effect took place. From today’s point of view, I have to admit that maybe we overestimated everyone being so annoyed that they would go to the dealer and get it fixed. But the experience showed that most of the people who had those effects wanted to get it fixed.
Q58 Chair: May I jump in here? To look at the case of Mr Gurung that brought all this to our attention, he did have that annoyance of not being able to get into his car, and he did get it fixed, but it was not identified that there was an issue that could have stopped his car while it was driving along.
Klaus Kompass: Chair, Mr Gurung had been annoyed and he went to a repair shop. He did not go to a professional repair shop. They tried to fix the car, but the car was not fixed. We found that one of the screws of the B+ terminal was not fixed after the attempt to repair it, and an inspection mirror was found at the B+ terminal. Someone was trying to fix the problem, but not in a very professional way.
Q59 Chair: But it is not unreasonable for a driver who identifies the annoyance and knows there is something wrong with his car to take it to a garage to get it fixed. In this instance, they were not able to identify the correct fix. Surely people are not always going to go to the dealership but to someone they think is a suitably qualified car mechanic.
Graeme Grieve: If I may answer, Chair, I have mentioned a number of learnings in this afternoon’s session, including prior warning. A clear lesson for us—and it should be a clear lesson for the industry and the DVSA—is that we must be more rigorous in how we apply prior warning and our analysis of prior warning. We took a view that was very much based on an engineering analysis and the impact that would have on our consumer. Clearly, a number of drivers did not respond to those prior alerts and prior warnings and the need to have the vehicle inspected. Our analysis on whether it was a safety-critical item was based on two factors: first, the RAPEX assessment; and secondly, prior warning. The prior warning is specific to the UK code, and that is why I believe that the Committee and the DVSA should review the interpretation and the use of prior warning, because in this case it clearly has not been understood or interpreted by a number of drivers—a limited number of drivers, but an important number none the less.
Chair: It is also about the action people take on the prior warning. Someone identifies there is a problem and they call the AA, who say, “You have to wiggle the cables about a bit and that will fix the problem”, which clearly it did not. They could call the AA, they could take it into their local garage, which is not a dealership with access to all your information, or they could take it to the dealership. It seems to me that there is a range of possibilities where that will not result in the correct action. Would you agree?
Graeme Grieve: I would agree. One of the key learnings here is that we need to be more careful in how we, as a manufacturer, apply prior warning going forward. There are also learnings here for the industry.
Chair: I think we now want to look at how you have managed the recall.
Q60 Jack Brereton: First, about the B+ connector recall, how many of the cars and vehicles that have been recalled do you think still need to be fixed?
Graeme Grieve: For the current recall, which we launched with first notification to drivers on 6 June, we have already repaired 30,000 vehicles, so we are currently running at a rate of around 10,000 per week. It is a No. 1 priority for me to complete that recall as quickly as possible. We have increased the number of customer service personnel that we have to deal with customer inquiries. We have mobilised our approved BMW retailer network. We have around 150 points around the UK, for which this will be a No. 1 priority. We are looking at quickly implementing a range of mobile BMW-approved technicians, which will allow us to complete the repair at people’s convenience, at their home or place of work. We are also carefully evaluating establishing some national dedicated facilities to allow us to complete this recall.
Q61 Jack Brereton: When do you think that will be completed?
Graeme Grieve: The recall will take until 2019 to be completed. We are working hard at the moment with our parts and components suppliers, because this requires replacement parts, to install additional production that is required to speed up this process.
Q62 Jack Brereton: Do you expect 100% of all the vehicles that have been affected by this issue to have been repaired and fixed?
Graeme Grieve: No, I do not. Based on our previous experience—I’m sure the DSVA will come back to second our evidence—we would expect to achieve a completion rate of 80% or more. Within the DVLA data, there will still be cars that have been scrapped, taken off the road or exported. It is unlikely that we will achieve 100% completion based on that DVLA dataset.
Q63 Jack Brereton: What are you doing to try to get to that extra 20% that you have identified as more challenging to reach?
Graeme Grieve: A few things. First, for those that are part of the DVLA data, there are two subsequent mailings if they do not respond to the recall, so that people we are trying to contact will receive at least three communications from us. Details of the recall are accessible from the front page of BMW.co.uk, which provides additional information. We have been very active on our Twitter feed, to provide data. We have also, as I am sure you are aware, had some recent coverage on the national programme “Watchdog”. We have been very proactive in our press-media communication. This issue has already achieved quite a lot of national coverage. I believe the awareness is very high. The current process will achieve a completion rate that is—
Q64 Jack Brereton: Have you looked at other potential data that you might use to locate some of those vehicles and locate addresses, for example if a car had been transferred to another owner or if somebody had moved? Have you looked at things such as MOT records or other types of data that might be held, which would enable you to locate those that are more challenging to find?
Graeme Grieve: At the moment, the ability we have is to access—via DVSA—DVLA data, which should take account of those driver changes. We do not have access to other sources of information other than that held by our own retailers. The approach to customers is getting more difficult following the GDPR regulations. Even if we do hold details of owners of vehicles who are no longer coming to us for service, they will almost certainly have not re-consented. The DVLA data is our most reliable dataset.
Q65 Jack Brereton: Do you have some customers who have just not responded at all?
Graeme Grieve: Bear in mind that the recall notification only went on 6 June. That will grow. We will then be re-mailing, and normally we would do it on a four to six-month progressive basis, if those drivers have not responded.
Q66 Jack Brereton: Is this the same for the blower regulator fix? Do you have these sort of situations? What proportion would be in the same situation?
Graeme Grieve: It is a similar date range of vehicles, so we would expect it to be a similar completion rate. What we do, together with the DVSA, is provide them with a rate of completion and certain milestones. They are agreed with the DVSA. If there is additional information that could be made available to us, we would clearly appreciate that, but at the moment, the industry works on the basis of the DVSA giving permission for the DVLA to provide manufacturers with current driver details.
Chair: We will now deal with the provision of information to the DVSA.
Q67 Steve Double: The DVSA say that in February 2016, they provided you with information about 21 B+ connector incidents and a further nine cases during 2016, and allege that you did not tell them about any cases during that time. Is it a fact that you really had no new incidents to report during that time?
Graeme Grieve: I do not believe that information is correct; it certainly does not tie up with our data. As I mentioned earlier, the first discussions were regarding a vehicle safety defect report in October 2014. At that stage, we did not interrogate our historical warranty records of previous cases. We did that only in preparation for a second meeting with the DVSA in February 2016, where we provided 20 warranty cases where the customer had indicated that there had been a loss of power. One of those was known to the DVSA, but the other 19 were not. We provided further information leading up to the decision in June 2016 to get the DVSA to halt that monitoring period. I am not sure as to what you are referring.
Q68 Steve Double: So you are saying that you told them about 20 cases in February 2016. What happened after that? Were there further cases after February 2016?
Graeme Grieve: There were an additional five cases from that meeting in February 2016 to the point in June 2016 when the monitoring period was ended and the review was closed by the DVSA.
Q69 Steve Double: As far as we are aware, they are saying no further cases were reported to them. What information have you provided to the DVSA to date—altogether? How many cases have you referred to them?
Graeme Grieve: I do not have a total number of cases. As I say, we are aware of 106 instances where the vehicle lost power, but there may be more cases that we have referred.
Q70 Steve Double: So it would be more than that, not less.
Graeme Grieve: I apologise; I will need to write to the Committee with that information. I do not have it.
Q71 Steve Double: That would be helpful. As far as you are aware, have BMW provided the DVSA with all the information that they have requested?
Graeme Grieve: As far as I am aware, we have provided the DVSA with all the information that we are required to under the code.
Steve Double: We would be grateful if you could write to us with that figure, because it would be helpful.
Chair: We are going to deal with issues around the provision of information to the public. We have already touched on it, but there are a few specific points that we want to raise.
Q72 Luke Pollard: If I am a customer of BMW with one of these affected vehicles, what information have you provided to me in relation to this fault?
Graeme Grieve: If you are a customer of a vehicle affected by the B+ battery connector, you will have received a letter from a mailing on 6 June this year, which would have alerted you to the fact that your vehicle is subject to recall. It would also have given you some additional details of what some of those symptoms of prior alert or prior warning would be to raise your awareness. If you are a customer who is subject to the heater blower regulator, you will receive that communication from us—it is mailing at the end of this week. We only recently received the dataset from the DVLA.
Q73 Luke Pollard: What advice are you giving to people about whether they should drive the vehicle between now and getting the fault fixed?
Graeme Grieve: With the B+ connector, we are alerting drivers to some of those symptoms of prior alert, which would be a very good indication of whether the vehicle was starting to suffer any degradation of the contact. Those drivers would either contact us directly on a customer information line or they would contact their local retailer for a booking to have the car inspected and repaired.
Q74 Luke Pollard: You are not saying, “In the event of there being any flickering, do not drive it. We will come and collect it.” You are just saying, “Book it in as soon as possible.”
Graeme Grieve: We are advising them of the signs of prior warning. Clearly, our objective is to repair all these vehicles as quickly as possible, but we would clearly prioritise any that are showing any of those prior alerts or signs of prior warning.
Q75 Luke Pollard: How are you conveying the severity of the problem in your communications? Are you just saying, “A fault has been identified and you need to get it fixed,” or is there something that says there is an urgency about this?
Graeme Grieve: I believe that the wording of the recall letter gives the driver sufficient urgency that there is an item on their vehicle that needs attention. The fact that we have included signs of prior warning would also give them an indication. If they are not having any of those signs of prior warning, there is less immediacy. If they are, there clearly is high immediacy.
Q76 Luke Pollard: How do customers react to getting correspondence like that?
Graeme Grieve: So far, customer reaction to the correspondence has been largely positive. We took a lot of calls following the “Watchdog” programme where customers reported concerns as to whether the car was affected or not—many were not. Since the recall letter, we have had a number of discussions with customers. We have had some customers who do not wish to drive their car. In those cases, we have offered to collect their vehicle and have it taken to a BMW service centre to await repair.
Q77 Luke Pollard: Earlier you mentioned garages subscribing to a system that gives them access to safety information and what to do about it. Do you know how many garages in the UK are subscribed to that system that are non-BMW main dealers?
Graeme Grieve: We have 880 subscribers. To clarify what the system is and what it does, we call it our after-sales online system. That was launched in 2016. We had a version of the same system from 2002. This is a system that provides information on technical campaigns and diagnostic data. If it is a BMW specialist and they want to use enhanced functionality, it would give them some ability to re-programme vehicles as well as diagnose. Broadly, it is a garage diagnostic and information tool. Information relating to safety campaigns or recalls is available to any independent via the DVSA website by chassis code search.
Q78 Luke Pollard: So safety-critical information is not being denied to garages so that they have to pay a fee. That is readily available to garages, whether BMW or otherwise. Is that right?
Graeme Grieve: Any garage or individual can use the DVSA website by searching chassis numbers to understand whether there are any outstanding recalls on the vehicle. I know from previous Committee sittings that the DVSA is looking to expand the use of that to potentially an MOT check—
Q79 Luke Pollard: Instead of looking at recalls, which are sometimes accompanied by large amounts of media coverage, let’s look at the B+ connector problem. Would that information have been available freely to garages during this five-year period where you were identifying problems, even though at some stages you said they were not safety critical, or would that information have only been available to those garages that were subscribing and paying to access the system?
Graeme Grieve: The difference will be that the first technical campaign on the B+ connector, which was at the end of 2016, would have been visible on the after-sales online system. As it was a technical campaign and not a safety recall, it would not have been present on the DVSA look-up site. At the moment, as I mentioned earlier, what is lacking is a cross-industry platform to provide information on technical campaigns. There is only one where there are recalls, which are safety critical. We as a manufacturer provide a system. A number of other manufacturers also provide systems. A number of these independents will specialise in one particular brand, whether it is BMW or another brand, and therefore are very likely to use the system, but if an independent is maybe working on 20 or 30 different car brands, they are unlikely to use a multiplicity of systems.
Q80 Luke Pollard: To help me understand it, until this became safety critical, independent garages that may have had someone turn up with one of your vehicles experiencing difficulties would not have been able to find out about the B+ connector problem unless they subscribed to your system. Is that correct?
Graeme Grieve: Correct. Unless they were a subscriber, they would not have been able to go through the diagnostic process that would have alerted them to that being a possible cause.
Q81 Luke Pollard: Do you think that is the right system to adopt? To go back to your statement about learnings, it seems that, from my point of view, the problems could have been caught earlier at various stages. Is denying independent garages access to that information something that you have included as part of your learnings or are you still comfortable that independent garages will not get access to those kind of technical problems that could be safety critical in the future unless they pay you money?
Graeme Grieve: First, Mr Pollard, I would like to go back and explain some of the terminology I have used, because it may have caused some confusion, for which I apologise. Up until the B+ connector being a technical campaign at the end of 2016, that information would not have been visible as a technical campaign even to our own retailers. Before that, the prior level of information would be a diagnostic tool—a diagnostic process. That is what was launched in July 2010 that alerted our retailers that if a driver is experiencing issues with central locking not starting, one of the items to check would be a B+ connector.
Q82 Luke Pollard: So independent garages have not been denied access to any information that would have enabled them to have spotted this problem earlier because they were not necessarily subscribers to your system.
Graeme Grieve: Do you mind repeating the question?
Q83 Luke Pollard: I am trying to understand whether any independent garages that may have had someone turn up with one of your vehicles with this problem have been denied information that would have helped them to identify the problem and fix it because they were not subscribing to your system.
Graeme Grieve: First, I do not think that “denied” would be the term I would use. There are independent garages that are not subscribing to our after-sales online tool that will not have some of the diagnostic functionality that that tool provides. Where it is related to a safety-related campaign in a recall, which is the key topic we are talking about here, the data on the car being affected would be available from the DVSA website. That would not necessarily provide them with details on how the repair should be carried out, but it would alert the independent that there was an outstanding recall on that vehicle.
I am not sure if that helps. Initially, before the technical campaign, this was an issue that was included in our diagnostic process. We receive literally hundreds of PuMA updates—technical update requests—every month in terms of codes that say to a technician, “If your driver is experiencing a problem, these are the things you should look at.” It is a diagnostic process that goes behind. Clearly that is across all areas of the vehicles; that is not safety. Safety would be specifically identified as part of a recall.
Q84 Luke Pollard: How have you told independent dealers and garages about this particular problem and what to do about it? Is it through direct correspondence or is it only if they have accessed the system or put in the chassis number, as you said?
Graeme Grieve: If we are talking about the recall part of it, which I think is the most impactful, we would not wish an independent to carry out that repair work; we would wish that to be done by a BMW-approved technician using BMW-approved parts. If a driver took their recall letter to an independent garage, that independent garage should refer them to a BMW retailer, because they do not have the capacity to do that repair themselves.
Q85 Luke Pollard: Great. For clarity, if someone took one of your vehicles, potentially with that problem, to a BMW garage, there would be no charge to the consumer for that work?
Graeme Grieve: Absolutely. We fully recognise the inconvenience that this is causing our drivers, and that repair would be conducted as quickly as we could, at no cost to the customer. We will try to do everything to minimise the inconvenience.
Q86 Luke Pollard: Do you think that the communications that you have sent out, and are sending out on the other problem you mentioned, are sufficient to get the vast majority of owners going to a BMW garage to get these problems sorted?
Graeme Grieve: I believe so, with the actions we have taken on our own website and via social media. We rely on the quality of data we expect from the DVLA. Also bearing in mind the very high media profile on these recalls, yes, I would expect drivers to be aware.
Q87 Luke Pollard: At what point does your recall activity effectively, for this activity, cease? Are you planning this correspondence up to a certain point? If someone does not reply to their first letter do you send them a second one, for instance? At what stage would you, as a business, be satisfied that you have done what you need to do to get the vast majority of people?
Graeme Grieve: Again, this is industry practice. There would be three mailings. The second mailing would obviously go to customers who have not had the repair conducted, and again the third to those who had not had the repair conducted. When we send those depends to a certain extent on the completion rate, but if we look at previous campaigns they would normally be interspersed by four to six months. At that point, we have to reach a conclusion with the DVSA as to whether we have done everything we can to complete that recall.
In answer to the earlier question, it is unlikely we will ever achieve 100% recall, particularly with the age range of cars, because many of these will no longer be in use, some could have been exported to other right-hand-drive markets, and some could simply be off the road.
Q88 Luke Pollard: That is the case in all EU countries where you have identified the problem—you are having similar recalling activities for these vehicles in every EU country?
Graeme Grieve: In terms of the customer notification process? I cannot answer for other EU countries; maybe Johann can.
Johann Ebenbichler (Translation): For all technical campaigns we check the recall quota on a general level as an average, but we also check it on a country-specific basis. We do it once a month within BMW AG, and if we see that one country has a lower rate than another, we talk to that country and ask what the problem is and why their rate is not higher. To give you an example, the recall for the B+ of January 2013, which was then rolled out, now covers 800,000 vehicles and the success rate is now 92%, which is very high, bearing in mind that older vehicles may have been scrapped, are no longer live or have been exported to a third-party country.
We see that, and we now have experience figures for those vehicles, so we are now familiar with the normal return rate for a 10-year-old car. If the figure is below that rate, we carry out a further action.
Klaus Kompass: As a member of the research and development team at BMW, in the majority of my work I am looking forward to new vehicles. I am responsible for vehicle safety and have worked in this field for 30 years now. We are working a lot on improving vehicle safety. As part of the work we do for future vehicles, we have identified that the recall completion rate should be higher and we want to do as much as we can to increase that.
We are working on a system that will go live next year to use BMW ConnectedDrive and the BMW Connected app to directly inform customers, sending a message into the vehicle and on to the smartphone, when we feel that it is urgent enough for them to go to the repair shop. That will probably increase the completion rate dramatically. Of course, we need the technical background for that, so that will not affect those cars but only new cars.
Q89 Chair: A couple of final points. When you extended the recall on the day of the “Watchdog” programme, how did that relate to your risk assessment process?
Graeme Grieve: When we extended the B+ connector recall, the DVSA were notified the same day as the “Watchdog” programme. That is not ideally what we wish to do, but clearly “Watchdog” had generated a significant amount of concern, particularly among our drivers, and we wanted to try to address that as quickly as possible. On the heating blower regulator, the DVSA were informed the previous week to the “Watchdog” programme in a meeting we had on 18 May that there would be a recall for that component.
Q90 Chair: Would there have been that recall if “Watchdog” had not featured it on its programme?
Graeme Grieve: Yes, there would have been. Both of these issues were clearly under detailed evaluation and at the point of conclusion. The timing with the “Watchdog” programme did generate a significant amount of media coverage and a lot of consumer concern—we totally appreciate and understood that—which, if we had been able to communicate earlier, we may have been able to dissipate to some extent. But the recalls were not called because of the “Watchdog” programme; that programme alerted consumers to these two issues but the decision-making process was independent.
Q91 Chair: So the timing was coincidental and the decision to extend that recall was down to your own risk assessments, not down to the fact that it was about to feature on the “Watchdog” programme?
Graeme Grieve: No. The “Watchdog” programme clearly put additional pressure on, in terms of the time for action, but it didn’t influence the decision. The decisions were already taken; we just needed to implement them more quickly—by which I mean a matter of days—than we otherwise would have.
Q92 Chair: Do members of the Committee have any further questions? Mr Grieve, can I ask you, as a bit of homework, if you would mind summarising for us the learning points that BMW has taken from this whole episode? Could you also set out where BMW thinks an industry-wide solution is required and where a change in process is needed? Can you write to us with that submission after the session? We may also write to you with our own further questions. That concludes our first session. Thank you very much for attending.
Examination of witness
Q93 Chair: Thank you very much for joining us. I am sorry that we are running a little late. For the benefit of our record, will you please introduce yourself?
Gareth Llewellyn: I am Gareth Llewellyn and I am the chief executive of the Driver and Vehicle Standards Agency.
Chair: I think we will start with some questions about the B+ connector recall. Jack will set us off.
Q94 Jack Brereton: Can you please first go into a bit of detail about BMW’s being late in reporting the issue to the DVSA?
Gareth Llewellyn: We first knew about this on, I think, 30 October 2014, when a member of the public contacted us. She said that she had already contacted BMW to say that she had a problem with her BMW 3 Series, which had cut out while in motion. That was reported to us on 30 October. As we normally do, we then wrote to the manufacturer—BMW in this case—to say that there was a particular issue. A standard set of questions then follows that, one of which is how many other reports they had of similar incidents. That went off on 31 October 2014.
We didn’t get a response, so we sent a chaser on 21 January. That letter is quite important, because it said that we had also heard of reports of vehicles cutting out that were outside of the original 2013 recall. We asked whether they would consider extending that recall, and if not, why not.
We then received a response on 28 January 2015 that gave us some information, but not the information we were looking for. We had asked for the number of reports but we did not get that. We just got the number of warranty claims that had been made, of which we think there were 14 where the vehicles cut out but five where they cut out in motion. There was a line at the bottom saying that BMW didn’t consider those to be safety critical. We believed that there was an issue, which is obviously why we flagged it to the manufacturer. The information that we had back wasn’t sufficient for us to make a judgment on a recall at that time.
Q95 Jack Brereton: You suggested that you believe some of the information to be inaccurate. What was inaccurate about it?
Gareth Llewellyn: I can only manage the recall process if I have the information on the root cause and what the defects are and so on. Being told that there are simply a number of warranty claims doesn’t give me enough information to judge whether that is a really serious issue, a minor issue—
Q96 Jack Brereton: Did BMW not provide all the information that you requested?
Gareth Llewellyn: No, they didn’t. We have asked on a number of occasions for that information and we have still yet to receive it.
Q97 Jack Brereton: BMW have denied providing inaccurate information. How have you responded to that, and what actions have you taken?
Gareth Llewellyn: All the way through this process we have asked two fundamental questions. One is how many more incidents of that type they are aware of. A number of figures have been quoted in the prior evidence that are completely new to me. I mean, they referred to 106 cases being referred to us. I’ve no idea where they came from. As I said, back in early 2015, we knew of 14 plus five warranty cases, but we didn’t know of any more than that, and we provided back to them—I think—five vehicle safety defects in 2014. I think there were 13 in 2015, and a further 15 in 2016.
Q98 Jack Brereton: BMW, in their written evidence, have said that DVSA did not provide clarification around these points.
Gareth Llewellyn: We constantly asked in meetings and in minutes, “Can we have more information? Can you give us any more reports or information that you have?” To be fair, we manage a lot of recalls with BMW. I think that 44 are in progress at the moment. The vast majority of them work really well, so this is not a very unusual relationship; it’s something that we have with most manufacturers. We’re aware of the sorts of information that flow between manufacturers, and us as the regulator, to enable us to make those decisions. It just felt, in this case, like the flow wasn’t quite as it was with many other recalls.
Q99 Jack Brereton: The recall process changed, because in February 2017 your mind changed about pressing for that recall. Why did you change your mind at that point and decide that it was necessary to extend what they were doing?
Gareth Llewellyn: As a general view of the overall recall process, if I can just go back a little bit in history—and I have been here twice before on the Vauxhall issue—I joined DVSA just over two years ago. It’s true to say that at the time, the culture around the recall process was very much embedded in the code of practice. There was a much closer working relationship with manufacturers, and a sense that if manufacturers didn’t call their own safety recall, we almost had to live with it, because we were managing a code of practice.
I’ve said at previous TSC inquiries about Vauxhall that when I joined DVSA, I got very frustrated in the context of the Zafira fires that there was little I could do. We formed a very productive relationship with trading standards in Luton, because they have the powers to enter, search and seize, and also have some of the powers that, at the time, we believed we didn’t. We have gone all the way through that process, and it’s a very complicated investigation. We’re currently sifting through 1.5 million documents as part of that investigation, so it’s a very sizeable operation for us.
The reason why I’m telling you that story is that as we went through that journey, we started to question whether our view of our powers was correct, and whether the strength of the regulator was as weak as it might have been in 2014-15. I sought legal counsel as to whether, in the context of the General Product Safety Regulations—where we’re actually not named as an enforcement agency, but which still apply to us—our powers were as weak as we might have thought they were. Actually, counsel have now come back with a very different view, so we do have the powers to enforce a safety recall. We have the powers to take a criminal prosecution. We have the powers—
Q100 Jack Brereton: I think that we will come on to powers a bit more later on. In the case of the unfortunate death of Mr Gurung, why did you not take action immediately once that evidence was presented?
Gareth Llewellyn: I still think that the view that we had back from the manufacturer was that, under their RAPEX risk assessment, there wasn’t a serious risk. If you have a look at the RAPEX risk assessment, it starts from an interesting perspective, which is, “How can you assess the risk of an engine stalling and a person being injured as a result of the incident?” But we know now that it’s not the engine stalling that’s the issue; the issue is that you’ve lost all electric power.
I think that in the run-up to that fatal accident, we had come to the conclusion that there were more incidents of vehicles stopping while in motion without any prior warning, and that’s the point at which we knew to make a change.
Q101 Jack Brereton: The coroner said, “‘I find no action had been taken by the DVSA to escalate the requirement for a recall of all potentially affected vehicles with BMW UK” between February 2017 and the start of the inquest. Why is it that the DVSA did not request that further action be taken at that time?
Gareth Llewellyn: We posed that question back as early as 21 January 2015. As I think I mentioned earlier on, we had said to BMW, “We’re now aware of more incidents with vehicles that are outside the original recall of 2013—that is, those that had the hot climate pack. Should you extend the recall, and if not, why not?” Just bear in mind that that is against the culture that I explained at the start, where there was much greater reliance on the manufacturers doing the right thing in terms of the recall.
We had been asking that question all the way through. It’s not surprising with hindsight that when you have an accident of that type, you have to decide that you should be taking the recall in a different direction.
Q102 Jack Brereton: Did you disagree or agree with the scope of the initial recall?
Gareth Llewellyn: In hindsight, I would disagree with the scope of the initial recall.
Q103 Jack Brereton: Not with hindsight. At the time, did you agree or disagree with the scope of the recall?
Gareth Llewellyn: The recall in 2013?
Q104 Jack Brereton: The initial recall that was taken in 2013.
Gareth Llewellyn: The initial recall in 2013 was only for—there is a bit of argument about whether it was for 78 or 87 vehicles, but it was a very small number.
Q105 Jack Brereton: And in February 2017?
Gareth Llewellyn: In February 2017, we had changed out people within the vehicle safety branch, and there was a much more robust dialogue with BMW around whether there should be a very clear safety recall, covering not just petrol engines but diesel engines. As with all regulators, I guess you regret not taking that decision much earlier.
Q106 Jack Brereton: So what action did you take at that time to say to BMW that the recall needed to be much wider and to extend the recall?
Gareth Llewellyn: We had those conversations. They were minuted and they went back and forwards in terms of emails. At that point, we did not believe that we had anywhere to go in terms of forcing the manufacturer to undertake the recall. We had that advice only in May this year.
Q107 Steve Double: If I heard you rightly, you said that there are 44 recalls for BMW at the moment.
Gareth Llewellyn: Yes, that is right.
Q108 Steve Double: For comparison, how typical is that? How does that compare with other manufacturers?
Gareth Llewellyn: Overall, last year, we had 403 new safety recalls, which is a reasonably large number. I think we managed something like 1,500 over the last five years, so that tells you that last year was a slightly bigger year than those. I would not assume that a large number of recalls is a bad thing. We encourage manufacturers to tell us early about defects, to recall them and to get them fixed very quickly. There are some manufacturers out there with over 100 recalls, but actually that is a really good thing, because we work very proactively with them.
Q109 Steve Double: That is helpful to get the context of what we are talking about.
Gareth Llewellyn: Those 403 new recalls cover something like 7% of the entire fleet, so it is way over 2 million vehicles.
Q110 Chair: Can I come back to the question about when you made the decision and also about your record keeping? We heard earlier, and you heard the earlier evidence, that in February 2016, there was a meeting between BMW and the DVSA, and a comment was written down by a BMW engineer that obviously alluded to the fact that there was the potential for a fatality to arise as a result of this B+ connector fault. The suggestion from the BMW witnesses earlier was that that comment was made by someone in the DVSA staff. What records do you at the DVSA have of that meeting in February 2016?
Gareth Llewellyn: I cannot comment on whether that comment was made, not least because I was not in the organisation at the time and the person who is alleged to have made that comment also does not work for us. I cannot confirm whether that comment was made, I am afraid.
Q111 Chair: What records do you have of that meeting?
Gareth Llewellyn: I do not believe we have formal minutes of that meeting. If I can extend that, one of the learning points for me is that the administrative side of what we did in 2014-15 probably does fit in with the words the coroner used to describe the DVSA. The way in which the administrative aspects of the recall process were being managed at that time was not up to what we would expect to be professional standards.
Q112 Chair: There was an opportunity there to identify that this was a safety issue, and had that been identified then, Mr Gurung may still have been with us.
Gareth Llewellyn: That is very true. I should also add that, as I sit here today, our thoughts are with Mr Gurung’s wife and family. I have spent the vast majority of my career in safety roles all over the world, and when things such as this happen, it is heartbreaking and you try to learn as much as you can.
I go back, though, to the statement that we made on 21 January 2015 back to BMW, which was a chasing letter for the information that we were looking for, which was how many other incidents they had had up to the point at which we gave them evidence on 31 October. That letter made it very clear that, “We are now seeing incidents occurring with vehicles that were outside the original safety recall. Should you be including this in the recall? If not, why not?” We asked that very fundamental question at that time. That is documented. As poor as our administrative standards were at the time, we made it very clear that we believed there was a safety risk at that time.
Q113 Chair: But you did not follow it up sufficiently?
Gareth Llewellyn: We did not have the information to follow it up, that is very true.
Q114 Chair: You will have heard that in the earlier session there was quite a lot of discussion about prior warning. From a DVSA point of view, what constitutes a serious defect, and how do you define prior warning?
Gareth Llewellyn: Prior warning is interesting; it has to be clear and unambiguous, and it has to give the driver sufficient time that he can bring the car to a safe position.
I have a personal view on prior warning: I have sat in a car park on a bank holiday Monday, trying to get into a car. I could not physically get in. Two things spring to mind, “Has the battery in my remote control gone, or has the battery in the car gone?” I don’t think at that stage it would have even crossed my mind whether that was signalling that the vehicle would stop in motion. I do not claim to be the most intelligent person in the world, but I think most people in society would probably feel that way as well. Equally, if something flashes up on your dashboard and then disappears, you do not instantly think, “There’s a major problem with my vehicle.” I am not convinced that there is a prior warning for something where all electrics fail while you are on the move. I cannot see that.
If I could just add one other thing in that context, which again is a learning point for us, on the day of the accident probably the most valuable thing at the time would have been for the hazard warning lights to have worked. It is rather strange—I have experienced this in other safety-critical industries—but you normally have your safety systems as fail-safe. It is amazing that hazard warning lights are attached to the same terminal as everything else in the car so that when you lose power, you lose the one thing you need at that time. We understand that that is a manufacturing consistency across most operations, and I wonder whether that is a future modification that would have paved the way for a better outcome here.
Q115 Chair: Has the DVSA’s thinking about what constitutes prior warning shifted since the coroner reported her findings in the inquest into the death of Mr Gurung? If so, what are you doing on that?
Gareth Llewellyn: There are two answers. Yes, it certainly has shifted, because we now realise that there is a particular incident here where there is no potential for prior warning. Part of the learning for us out of this whole incident is to review the safety recall process, to try to understand at what point a defect simply fails to give enough notice to the driver that they can bring their car to a safe position. The threshold for safety recalls will be adjusted on that basis, but we are not in that position just at the moment.
Q116 Chair: It seems to me that it is not just about whether you can bring the car to a stop; it is what you do afterwards, as we have heard. For some people who had the problems, there was a prior warning—they had a problem and perhaps contacted the AA or took it into their local garage, and a fix was applied—but it did not tackle the underlying problem indicating that there was a defect that had quite serious safety implications. How will you deal with that in the future?
Gareth Llewellyn: One of the roles we should have with manufacturers is just scoping the initial risk assessment. As I have said before, in my mind the RAPEX risk assessment for the very first occasion focused on absolutely the wrong thing, which was the engine stalling. We critiqued that for BMW and went back to them and said, “That risk assessment needs to be on the total failure of your electrical system while you are on the move.” To be fair to BMW, they have highlighted something that is in the back of our mind—there needs to be a much earlier conversation on the risk before defining whether prior warning is a sensible conclusion in that space. It was a core action out of the Vauxhall issue as well. Vauxhall were clearly focused not on the root cause but on a symptom. We need to be able to get industry to a position where they focus much more on what they think the actual root cause is.
Q117 Chair: I am interested in the wider implications of this. Does that shift in thinking on the risk assessment process that you have just described have implications for some of the other cases that the DVSA has looked at, such as the Nissan Navara chassis problems—we have seen some pretty dramatic photos of cars that are literally broken in half—or the Ford Ecoboost coolant pipe failures? What does it mean for those cases?
Gareth Llewellyn: I would not put either of those in the safety recall process at this stage. With the Ecoboost there are three levels of warnings, so the vehicle does not fail catastrophically in one go. There is sufficient warning for the driver to take action in a timely manner. The Nissan Navara is a slightly different issue. We believe that that is not a design defect, but wear and tear over the lifetime of the vehicle. That is the sort of thing you would expect to be picked up in the MOT process.
Q118 Chair: So the reason why the BMW B+ connector warrants a recall but the others do not is that you think the high warnings are sufficient that someone will have taken action?
Gareth Llewellyn: That is certainly true in the Ecoboost case, because it happens in a very short period of time but you still have enough warning. The issue with the Nissan Navara is the extent to which that chassis is degrading over a number of MOT checks. That is your annual check effectively to see whether that risk is increasing or not.
Q119 Chair: Based on the work that you have done in relation to the Nissan case, from the photos I’ve seen it is not clear to me whether anyone’s car has fallen to bits while driving along. You are saying that it should not happen, that someone should have realised that the chassis was deteriorating sufficiently that they would have taken remedial action. Is that right?
Gareth Llewellyn: Certainly, when you have got it on a stand doing an MOT, you should be able to see whether it is degrading to that extent. I have also seen the pictures and you do wonder how they had got to that stage.
Q120 Chair: Finally, to clarify again, in February 2017, you changed your mind about the B+ connector. Was that the availability of new information or was it that reassessment of what was meant by a prior warning that caused you to think differently about the issue?
Gareth Llewellyn: It is the latter because, on 20 December 2016, that was the first occasion we had a defect where we knew that the vehicle had failed in motion for the first time without any prior warning. Up to then, there had always been the prior warning round, such as, “I couldn’t access the vehicle while it was on my drive,” and so on. But that was the very first occasion.
Q121 Chair: Turning to the blower regulator recall, in your written evidence you say that the DVSA is still awaiting official notification from BMW of its intention to recall affected cars. But appendix A ends saying, “23 May 2018 BMW advised the DVSA of its intention to recall cars potentially affected by the heater blower issue.” Are those two things different or are we missing something?
Gareth Llewellyn: They are different. The word is “intention” there. We received the official recall notice for the blower heater resistor this morning, and it has gone live on DVSA’s website today.
Q122 Chair: So, you knew they were intending on 23 May but you had not had some official notification.
Gareth Llewellyn: To be fair, that is quite an important distinction because, on that recall notification, there is all the information we need to be able to convey to the public and also to be able to start monitoring the progress of the manufacturer in meeting that recall.
On the B+ recall, we still do not have an official date, so it is on our website in narrative version rather than a particular start date, so I cannot monitor the progress. There are also other consequences of not having the start date on the system, which we can explore in terms of consumer information, but I cannot monitor the performance of the manufacturer until that start date is clear.
Q123 Chair: Let me be clear about the information that is provided and what has not been. BMW say you were informed at a meeting on 18 May, and that was confirmed in email correspondence on 23 May. Is that correct?
Gareth Llewellyn: That is true, with one exception. It still says the official launch date is to be confirmed.
Q124 Chair: Right. So you knew they were going to do a recall but you just did not know the date from which it was going to start. Is there anything other than that official start date that you need from BMW that you did not get at either the 18 May meeting or in the email of 23 May?
Gareth Llewellyn: No, it is just that start date. Although that might a technicality, could you give me a few moments to explain why that is so important? BMW are right that the greatest route they have is through the registered keeper information that we provide through DVLA.
One of the changes since I was last here is what we make available on the MOT history site. You can enter your registration number—not your chassis number, as was mentioned a few minutes ago—and it will give you the entire MOT history of your vehicle. It also gives you the opportunity to check whether your vehicle is subject to a safety recall. That is very much a reactive process.
However, if you are thinking of buying a car and you go on to a website or go through a trade magazine, you quite often will not get the registration number. You will just be given a description of the vehicle. The only other route for you to know whether there was a safety recall outstanding for that is if you go to the list of recalls we publish on gov.uk. That only gets logged on there when it has a start date. There is a group of people out there who might be buying a car potentially not knowing whether it is subject to a safety recall.
Chair: I think we want to move on to ask further about DVSA’s management of recalls. Luke, I think you are going to pick up here.
Q125 Luke Pollard: In her report, the coroner noted a general lack of documentation underpinning the DVSA’s evidence, with “no detailed minutes of meetings taken; no notes of internal discussions; and nothing to record the procedure or any timetabling for action to review the reported faults”. What concerns do you have about the criticisms that the coroner made of the DVSA’s record keeping? They seem to be pretty serious. Can you tell the Committee what steps you have taken to address those concerns, please?
Gareth Llewellyn: I accept all those concerns. I think, at the time that we were in the early stages of this recall, they were absolutely right. The administrative side of the safety recall was not where it should have been. We have instituted a different process now. Part of it is driven by clarity over what our powers are, so there is a set series of documents as we go along through the process. The two that are most important, we have added, and I will explain the timescales in a minute. The first one is what I call a “minded to” letter. It is a letter to a manufacturer saying, “I am minded to force you to issue a recall notice, because I have the legal powers to do so. You have 10 days to tell me why I should not do that.” In that notification, we make it clear that if we do publish a recall notice, that will be made public, so if you cannot convince me that you have given me all the information as to why there should not be a recall, I will force you to undertake a recall, and I will make that public. There is a reputation and a brand issue in all that, of course. But it also then triggers a whole range of other things that I can do. I can prosecute. I can actually go and inform all the customers with that particular vehicle that I have asked or told the manufacturer to undertake a recall. So the process towards the end of it is much, much more robust now.
On the administrative side, we have added casework officers and administrative officers into the recall team, to make sure all the data and record keeping are up to scratch, because I am really conscious that when we come here—I hope I never have to come here again—the audit trail of the decisions that have been made should be absolutely robust for your benefit.
Q126 Luke Pollard: I realise that you were not in charge of the organisation when large parts of this activity were going on. When you first took over the role, were you aware that there were these issues with the DVSA, or is it something that you have had to become aware of, or you have become aware of, during your tenure?
Gareth Llewellyn: It is certainly the latter. I am not short of any challenges in my role. It is true to say that our enforcement activities, of which this is one, were embedded within our core operation, so they were quite low down in the organisation. One of the decisions I took—I think it was on 31 March 2017—was to take all our enforcement activities out into a separate branch, reporting directly to me and sitting on the executive, to try to give it the profile and also the rigour that was necessary. We are working our way through that process. How can we make sure that the resources are right? How can we make sure that the administrative processes are right and also that, when things go wrong, we can step in and make things better, as we have done with the Vauxhall investigation?
Q127 Luke Pollard: Can I just ask about DVSA powers? Something you said earlier surprised me. You said you effectively, as an organisation, did not know that you had certain powers. I am just trying to understand: how does an organisation not know that it has enforcement powers? That seems to be one of the fundamental basics of your organisation. Could you explain how you did not know?
Gareth Llewellyn: I just want to give you reassurance that it is not universal across my organisation. We had 66—sorry, 69—successful prosecutions last year for Bluetooth fraud in the theory test. We found 28,000 defects on HGVs and buses around the country, some of which would match the photographs you have seen—on the Nissan Navara, for example. So, for the vast majority of cases in my enforcement branch, they do an absolutely fantastic job and we are safer because of their work.
It is interesting that when you look at the general product safety regulations, that references local authorities as having powers but doesn’t reference us, so we were being asked to implement the content of the regulations through a code of practice, but it was absolutely silent on what our powers were. I think the legal counsel has come to the view that, actually, there were legal powers; it is just they weren’t explicitly set out in words that said, “DVSA is an enforcement authority.”
Q128 Luke Pollard: So you’re working at the moment on the basis of legal counsel advice that says you have these powers, even though the legislation doesn’t say you do.
Gareth Llewellyn: It does not explicitly name us. That’s the issue.
Q129 Luke Pollard: Is that something that you have raised with the Department, and what is their perspective on the decision you have taken?
Gareth Llewellyn: The Department are really supportive of our work in this area. They understand how difficult it is. I have absolutely no concerns about the support we get, from the permanent secretary to the director general. Indeed, the permanent secretary is on one of our enforcement sites on Friday to understand a little bit more about what we do.
Q130 Luke Pollard: So would you recommend—I’m thinking explicitly about the role of the Select Committee to scrutinise the work of the Department for Transport—that we suggest to the Department that they need to be explicit so that you are not operating under a presumption that you have powers, and you can point to the actual words in a piece of legislation, be it secondary or primary, that explicitly affords you powers in this respect?
Gareth Llewellyn: I think I’m content with where we are now. We are playing out a number of the “minded to” letters at the moment. The first one we sent out had an instant effect. The manufacturer instantly had its own recall, which is exactly the type of conversation that we want to push more broadly. We have got one that is probably due in the next couple of days, and it will be interesting to see whether that manufacturer treats it with exactly the same level of urgency that the first one did. So I am reasonably content with where we are.
There is an interesting quirk that still affects the Vauxhall investigation. I think the Consumer Rights Act 2015 failed to reinforce the role for DVSA in entry, search and seize. That is the only bit that we do not have access to. Interestingly, in the Vauxhall investigation, where we have used trading standards from Luton to achieve exactly that outcome, we would still need to do it under the current advice that we have. I should say that the trading standards from Luton have been absolutely fantastic in that process. It has been a real joy to work with them.
Q131 Luke Pollard: Could you write to the Committee about the technical detail of what you have just said, because I realise that is probably a level of information—
Gareth Llewellyn: Yes. Indeed, I sent a letter to the Chair explaining the difference in terms of where we thought we were back in 2014 and where we are now, just to be reassured about the Vauxhall investigation.
Q132 Luke Pollard: You have mentioned Luton trading standards a couple of times. I imagine that is because that is the home authority for vehicles. Do they have the responsibility for the entire country as the home authority?
Gareth Llewellyn: No, they just happen to be closest to Vauxhall’s head office.
Q133 Luke Pollard: Oh, right. Is there a single trading standards organisation that has responsibility? Some sectors—travel and tourism, for instance—have one trading standards across the country that takes the lead on certain matters.
Gareth Llewellyn: I cannot exactly answer that, but, thinking back to the conversation we had about how we exercise some powers in relation to Vauxhall, I think we probably would have gone for a national body if there had been one available. I will need to check that.
Q134 Luke Pollard: Great. In terms of the investigation around Vauxhall, is there any realistic prospect that Vauxhall will face charges?
Gareth Llewellyn: It is difficult to prejudge where we are at the moment. As I said, we have slimmed it down from something like 1.5 million to a third of a million documents that we are going through at the moment. Some of those are legally privileged, so we are having to take legal advice to sift out the things that we should not be looking at. I have to say that, right from the start, I felt there was more of a tone of withholding information from us as a regulator than there is with BMW. I think it was worse with Vauxhall, so I would be surprised if there is not some action that comes out of it, but time will tell.
Q135 Chair: May I just follow up on this issue around powers? The reason you have the memorandum of understanding with Luton trading standards is because that happened to be close to Vauxhall’s headquarters. I am not quite clear about your clarification. You think you know you do have enforcement powers. Does that mean that if there were an issue with another manufacturer based in a different local authority area, you would not have to establish a similar memorandum of understanding with another trading standards body? Is that right? That would not be necessary, or would it be?
Gareth Llewellyn: I’d hope not. If we have 403 recalls every year, that would be a very big overhead. I think it is unlikely. The clarity over the powers that we now have is really good. To be able to say to a manufacturer, “If you don’t give us this information, we will force you to undertake a recall, and if you do not do that, we will prosecute you,” is enough of an incentive. To be fair to most manufacturers, in the vast majority that we deal with on recalls, it works really well, so we are not here for 403 occasions during the year. BMW has managed 42 recalls with us, and they have gone rather smoothly. This is the backstop, but we want to be clear there is reputational damage if I force you as a regulator to undertake a safety recall on what can be quite a large fleet.
Chair: Finally, we want to ask some questions about the prosecution of owners.
Q136 Jack Brereton: As we have heard, there are potentially vehicles still on the road that should have undergone the process of recall and had repairs. We heard that potentially 20% are difficult to identify and find. Does the DVSA have any data on how many vehicles might still be on the road that do need these repairs to be carried out?
Gareth Llewellyn: I do not have that data at the moment. It would be relatively easy to pick it up through the MOT process because we obviously have all of the MOT data for each of the years and would know whether a vehicle is outstanding an MOT process, for example. Of course, once you take it into the garage for an MOT, that is the opportunity to check whether it has an outstanding recall.
Q137 Jack Brereton: Are you considering potentially collating that data at some point in the near future?
Gareth Llewellyn: It is something that had not crossed my mind until you mentioned it, but it is something worth doing, certainly.
Q138 Jack Brereton: As we also heard, there are concerns regarding drivers. Obviously drivers are becoming more informed. You yourselves have made more information available where there are recalls. Do you think it is appropriate that, where we have drivers who are ignoring this and not taking their cars in to have these fixes, they should be prosecuted?
Gareth Llewellyn: I think that is probably a policy decision for the Department. That is not me stepping back, but in Germany you get one opportunity to take it in for a recall. If you miss that one, on the second occasion my equivalent will take the number plates off your car and you will not be able to drive it. There is a different approach to the failure to take your vehicle in and get a safety recall remedied. We do not have that ability here.
Q139 Jack Brereton: So there are no examples of where that sort of action has been taken in the UK?
Gareth Llewellyn: Not here. It is interesting that if you were driving a HGV and we found you with a serious defect, we would immobilise it and you would not be able to drive it away.
Q140 Jack Brereton: So you would suggest that maybe there is a need for some legislative change there to improve the robustness of that area of policy?
Gareth Llewellyn: The first step is to analyse the data and see how much of a risk is out there from vehicles that have not had their safety recall undertaken. There is an interesting facet for us in our thinking about recalls in the future, because there is quite a wide range of safety recalls being undertaken, from the very serious one that we have discussed today through to the failure of the hydraulic strut holding your boot up. You would get a whack on the head from the second one, but probably not to the extent that it will be a fatal incident for many drivers. We will have to be a little more discerning about what sits in that recall category before deciding, if we go down the prosecution route, what cohort we target.
Q141 Jack Brereton: Who is currently responsible if we have an owner whose vehicle has been recalled but they have failed to take it back in to be fixed? Who is responsible currently, and who would face the consequences of that vehicle not being made road-legal?
Gareth Llewellyn: At the moment, under the general product safety regulations, all of those liabilities rest with the producer and distributor. They do not rest with the individual car owner.
Q142 Chair: Two questions remain. You heard the earlier evidence that was given by BMW and, in response to earlier questions in this session, you have said that they did not provide information or did not do it in a timely manner. Some of the public watching will just be thinking, “Why on earth is there this to and fro, or disagreement, about who provided what information when, when all the focus should be on keeping me safe on the road?” Why, in this case, does there seem to be a breakdown in communication between the DVSA and the manufacturer?
Gareth Llewellyn: As I have said, I think BMW have done a great job on 42 of the recalls. We have had a problem with these two in particular. Understanding what the prior warning is has obviously conditioned one side of the argument. When I looked at the RAPEX risk assessment they did, it clearly was not for the right risk. When we critiqued it, the second RAPEX risk assessment was different; it was all around electrical failure. Within there, there is a lack of understanding of what risk is, what the prior warning should be, and therefore the seriousness with which you treat this particular defect. Collectively, we have probably learned a lot out of that process. It is less about numbers in coming to the right decision. As you rightly point out, my role is to make sure that people stay safe on the roads. That is my role entirely. I can do that only if I have the right information. If I do not have information, I will be crystal ball-gazing a lot of the time. That is why that information is so important to me.
Q143 Chair: Have those lessons been learned, because obviously there has been an appearance in front of the previous Committee—albeit with a different manufacturer—around a safety issue. We would not want to be in the same position again.
Gareth Llewellyn: I certainly would not want to be in the same position again, and I do not think any manufacturer would like to be here either. The progression we have made from 2014 to now is quite significant. I sit here as a regulator knowing that I can prosecute a manufacturer for the first time if they do not launch a safety recall. When I first joined the organisation, I had none of that clarity on what our role was.
Q144 Chair: Finally, you were here and listened to the earlier evidence. Is there anything in particular that you wanted to comment on, which has not been covered in our questions?
Gareth Llewellyn: As I said, out of every safety incident there is a huge amount of learnings, which the organisation and myself will take. I would point out three things. A technical campaign is good when you can get access to the vehicles and you know who owns those vehicles. But where those vehicles are out of warranty and they have been sold on, there will always be a cohort of people that you cannot get to.
One of the learnings we will take back is whether, when we believe something is a safety defect and the manufacturer suggests a technical campaign, we say, “No, that has to be a recall,” because only then can we manage the cohort of people who sold their vehicle on. It is tragic in this particular case that we know that the owner of the BMW was not the first owner. They had bought the car five years beforehand, I think, so they would have been outside of the remit of this technical campaign—it would never have reached them, and that is rather sad. We need to reflect on that.
With the element around the design of the hazard lights, you want systems that are fail-safe. If he just had the hazard warning lights available on that day, we may be facing a quite different outcome. Why cars are designed so that when you lose all the electrics, you lose the one thing you absolutely need at the time, is something we will have to pick with colleagues in other agencies.
The other area for us, as we go through this process, is at what point do we make the lack of a safety recall an MOT failure. There is a lot complexity there around how we understand whether a vehicle has had its recall remedied, because there is a time lag between one garage doing it and then taking it into another garage for an MOT, so at what point do you make that in real time? There is also the question of being a little bit more discerning about what is a true, high-risk safety recall that has to be done, otherwise you will fail the MOT. There is work to do in all three of those areas.
Q145 Chair: What is your timeframe or target time for those issues to have been addressed?
Gareth Llewellyn: With the technical campaign, we can make that change to our code of practice when we have the next conversations with SMMT and its members. I think that is a relatively straightforward thing to do. The issue of hazard warning lights is out of my remit, so I put it out as a potential suggestion, because it feels, as an owner of a car, that I would not want my hazard warning lights to disappear.
Q146 Chair: What is the route for getting that issue raised and addressed?
Gareth Llewellyn: I will have to come back to you on that, because I am not entirely clear about who would make the final judgment that that change would have to be made. The MOT failure is something for us, so as we go through the next change in the evolution of the MOT guidance, we will be working with our policy colleagues to see how we can make that change.
Chair: We look forward to receiving updates from you on how those matters are being progressed. That concludes our session today. Thank you for attending.