Good afternoon, gentlemen. Thanks for coming. I have a number of questions. At the beginning of the Bill, we are told that automated vehicles are what the Secretary of State says are automated vehicles. Clearly, some thought has gone into the criteria for so-designated vehicles. What are your own thoughts and observations on how we can be sure we are getting that set of criteria correctQ ?
From an insurance perspective, that is one of the clauses we particularly welcomed when we saw the Bill. One of our concerns in advance was that it would not be clear to the customer what cars needed this new insurance system, so the clarification that the Government are going to take responsibility for doing that is really welcome. It means basically that we know what cars we need to have this new insurance for, and the customers will know that as well.
In terms of criteria, it is relatively simple. It is more about the user than the technology. I think the technology might not move that much, but it is the point where the user can feel confident that, when the car is in automated mode, it can deal with everything. Thinking about the worst-case scenario of an accident, if the car senses it is going to go into emergency mode, the car is able to do something to deal with that, which does not require the driver to come back in. We feel that if there is any point where the driver needs to come back in, it is not really an automated car. It is that tipping point where the car is completely capable of dealing with every situation. It might not carry on driving, but, at the very least, it would do an emergency stop and get you into a safe stop manoeuvre. That is the tipping point, or distinction, that we see.
Just to underline that, the measures in the Bill are designed to deal with the sorts of situations where a vehicle can drive itself in automated mode and not require the oversight of a human when the driving test is being operated. The particular mechanism by which those vehicles are going to be certified is an active topic of discussion at international regulatory forums. We have actively participated in those discussions, but we felt it was important to flag at the outset of the Bill that that would be clear to people in the insurance industry and elsewhere, to make sure they were able to understand which vehicles these measures apply to and which ones they do not.
In practice, we would be hoping that from an insurer’s perspective, it is pretty easy to find that out, just by looking at the licence plate or the VIN number. There would be a clear definition that this is a car with automated functionality.
Of course, we are now entering a major shift, because we have to have insurers who are going to be picking up the tab for accidents involving automated vehicles. Presumably, as we progress, we will be getting errorless driving in automated vehicles. That is the objective, and incidents will result. Therefore, we should be looking forward to cheaper insurance policies, but that may not automatically be the case, if you pardon the pun. How progressed or how ready is the insurance industry to deliver products that would make fully automated vehicles accessible to people in terms of costings, including the cost of insuranceQ ?
We are very advanced as an industry, particularly in the UK. Because of the clear message that the Government have given, we are perhaps ahead of our contemporaries in other countries. The two really important criteria in terms of the cost of insurance will be the volume of accidents. We are fairly confident—Thatcham Research has done quite a bit of research that suggests the number of accidents is going to come down a lot once we get automated driving. That will obviously reduce the number of insurance claims, which will inevitably have an impact on the cost of insurance.
One factor that we probably do not know about at this point is the actual cost of the vehicles themselves, and how much they cost to repair. We might have considerably fewer claims, but very high costs associated with repair might have an impact. That said, that is something that is happening already. Vehicle technology is changing a lot already, so it is not a case of a huge tipping point in technology once we switch to fully automated cars. The technological change will happen more steadily, so I am very confident that the insurance industry is ready to deliver competitive insurance products that will be affordable, will help people and will make them want to take up this technology.
Q Will you help me on another matter? When an accident is caused by an automated vehicle, we are told in the Bill that the insurer is liable for the damage, but when the automated vehicle is involved but is not insured, it is the owner of the vehicle who is responsible for the damage. I am wondering what your view is on whether that ought that to be the person in charge of the vehicle, rather than the owner. We might have the perverse situation of a stolen vehicle being involved in an accident but, according to this, the owner of the vehicle would be in the frame. Do you understand?
I think I know what clauses you are referring to and my understanding of them is that that covers publicly owned vehicles and Crown Estate vehicles. They would not have insurance because they do not need them. In those cases, where it is a publicly owned vehicle, the liability would fall on the public body. It is a separate arrangement for genuinely uninsured driving—private cars that are uninsured.
That is exactly right. That clause covers publicly owned vehicles. We anticipate the situation being similar to the situation at the moment for conventional vehicles. It is often the case that they self-insure, rather than going through an insurance company.
Q I do not know what your handle is on this, and that was terrific clarification, but do you not see that it is possible to interpret it in the way I did? A vehicle that has been stolen is not insured, but the owner of the vehicle is picking up the cost, not the person who stole it.
Q Finally from me, we are in the world of upgrades, which will present opportunities for manufacturers to continue to engage with the vehicle, in effect. I was wondering whether you had applied your minds to any future product liability exposures with the advent of new software. What does that mean in terms of those future liabilities and in terms of limitation, because current liability, once you part with the vehicle, has a 10-year limitation, although we have got some issues around extensions for people with disability insurance and so on? But if it is a product liability issue, there could be a succession of products that could give rise to liability. Is that factored into your thinking? Is it relevant? Have you dismissed it?
That is very relevant. When the consultation first came out, one of the questions was, “Do we bolt product liability into the motor insurance policy?” We looked at it in quite a lot of detail, and that was our initial assumption for how it would work. When we thought about it, those issues that you referred to and the fact that product liability only lasts for 10 years made it feel like too much of a change for product liability to be put directly into a Road Traffic Act situation. That is why we came to the conclusion that it should be a primary motor insurance policy, with the option then to recover from manufacturers.
Our conclusion is that you probably do not need to change the product liability, as it is kind of a backstop and it will not affect the original claimant. There might be a case on some occasions, if it is an older vehicle. We do not know quite how the market is going to develop—whether cars will be on the road for 10 years or longer in this situation, or maybe the product liability will renew itself every time there has been an upgrade. Let us assume that it does not, and you do expire at 10 years. My understanding is that there will still potentially be the option for a civil liability claim, so you might be able to argue that the manufacturer is not product liable, but they are in effect acting as the driver, so there is another claim that you might be able to bring.
Q You could be caught with a person with a disability, or a child, who is not subject to ordinary limitation; it would be from the date of their majority. There could be an action on the attaining of that majority and your product liability recompense from the manufacturer is effectively null and void.
Yes. I think that is factored in. The Bill means that that is a problem for the insurer, rather than the victim. I suppose part of the calculations that insurers will make is how many of those claims they will be likely to face. Are they insuring vehicles that are over 10 years old? That might have an impact. What is important in the Bill is that it makes that a problem for us as an industry. It will not affect, say, a disabled person who is using these vehicles. I think that is the insurance we need at the moment.
Q On this insurance issue, there has been quite a lot of speculation about what might happen to the products that the industry offers, which will clearly have to evolve. We were told that again this morning. Is it your estimation that that will affect premiums? One would expect premiums to fall, given the fact that these cars will be safer; many of your claims are related to human error, after all. Is that how you see things developing? However, we are also told that many people who cannot currently drive will now be able to—the infirm, the elderly, some disabled people. In a way, that is the most exciting thing about this development. How would that affect your assessment of premiums?
On the first case, I would think of it more in terms of claims costs than actually speculating on what the premiums would be. Obviously, if the number of accidents comes down dramatically, that is going to have a significant impact on the costs that insurers face. Motor insurance is very competitive, and it is inevitable that, if we see a significant reduction in costs, we will see a significant reduction in the premiums charged. So I think we can be pretty confident of that. As far as we know, it is still four or five years before these products will come to market.
Looking ahead to the cars you are referring to, where, say, there is a severely disabled person who possibly cannot drive at all at the moment, we are probably thinking about a level 5 car that can go from A to B in fully autonomous mode. It is fair to say that this legislation is primarily aimed at cars that will be manual for some of the time, automated for the rest: more of a level 4 car. Once you get to level 5, that is probably the point at which the insurance system is going to have to change more significantly.
Where the Bill is really helpful is that it allows us to learn from the first developments, get an insurance function in place and see that that system works. It is probable that we are going to have to evolve further once we get to a fully automated car. David Williams, who was one of your witnesses this morning, is one of the insurers involved in trials of fully automated technology. There is a significant degree of interest from insurers in the next generation of technology as well; but it is probably fair to say that this Bill is more around a level 4 car. I prefer to think of it as a binary distinction between automated and not-automated. I am not completely convinced about the levels and how useful they will be for consumers. It is probably fair to say that we think of it as level 4.
Q Public confidence in autonomous vehicles will be critical in terms of how quickly we can take advantage of some of the benefits that are portrayed by the industry. Given that the technology largely exists, do you feel that the Bill is going to go far enough to encourage the uptake of autonomous vehicles? I am specifically interested in whether you feel that the connectivity will allow a truly UK-wide uptake and also in rural areas, given that, as we have heard previously about electric vehicles, range and the ability to get to the destination is one of the limiting factors.
The point about confidence is really important. Trust in the technology is going to be a vital factor in seeing some of the benefits we are talking about. It is part of the reason that the Government are investing with industry in demonstration projects, which will involve members of the public trying the technology, understanding what it might mean for them and helping the developers to learn from that in terms of their public messaging and how they take the technology forward.
With regard to connectivity, what is interesting is that different developers are following different development paths for the technology, some of which rely on connectivity and some for which do not. So, from a Government perspective, it is difficult to say exactly what the final technological solution will look like. Some time is needed to work that through, but we are actively trialling this technology with the industrial players to understand, from a Government perspective, what action we need to take to make sure we are prepared for it.
Q Just to press that point about the end-to-end availability, is enough being done to ensure that you get truly wide coverage to allow people from outlying areas to get out and back again in the new technology?
It is an important bit of work that will to be done as we develop out the technology. We are investing in connected vehicle test-beds to understand what the requirement is, and, certainly, one aim of that work is to try to understand how it can benefit everyone, not just people in cities.
Q I would like to turn to clause 4, which relates to accidents resulting from unauthorised alterations or failure to update software. Hopefully, you have had a chance to look at this section of the Bill. It is all couched in terms of the operating system—interference with and failure to update the operating system. I am concerned that there are other aspects of software in a car that are relevant. Are you satisfied that the Bill is in the right shape, referring to the operating system, or would you prefer to see some other definition, in which case, what?
From our perspective, my initial reading of it was that it covered what we thought it was, and I am thinking it is the technology. I have say, I am not a software expert, so if the wording could be clearer—the clause basically says you are not liable if a stupid individual mucks around with the car’s systems and does things that the manufacturer would not permit.
Q We have a shared understanding of the purpose of this clause. I should say that I am a software engineer—or at least I was before I first came here—and I am slightly concerned that if this was tested in court and somebody had interfered or failed to update their firmware—a low level software—or if the driving was done by application software sitting on top of the operating system, the purpose of Bill, of which we have a shared understanding, could be defeated. You are saying that you have not formed a view on whether “operating system” is the right definition.
Q I have three quick points for Mr Howarth. First, the Bill talks about the vehicle being insured as opposed to what currently happens, from my understanding, which is that driver is insured. So I have policy motor insurance that enables me to drive certain vehicles, including my principal one. Is the insurance industry happy with what appears to be a change in focus—that is it now on the vehicle rather than the driver?
I think it is not a huge change in focus. In practice, the enforcement that industry currently does—via the motor insurance bureau—to check that you have insurance is done via the vehicle. It is done by checking licence plates. The responsibility is on the human driver, but the practical enforcement is to check whether that car, on the road, at that time, is covered by insurance. This Bill is primarily designed for vehicles that will be manually driven some of the time and automated some of the time. It is just the practicality that, once you are switching to an automated car, you need to be thinking about the car rather than the driver.
The whole purpose of the legislation is, I suppose, to be an enabler and say to people, “You can be confident using this technology, because you will not have to worry about getting into complex battles with your manufacturer.” We do not know for certain that that is what will happen. Some manufacturers have given positive statements about it, but if that does happen, the insurers will step into the front of the system and say, “We are actually going to take these on. We are the first port of call, even in the case where the person to whom we have sold the insurance policy to is not directly liable.”
Q The third point is on the software updates. From memory, when I bought my car, which has a touch screen on it for the radio and things, it would have cost me an extra £600 to have sat-nav put in. That is just the software because it already has the screen and the buttons and everything. I am thinking about software updates, which we have talked about, and a failure to install software updates could invalidate the insurance policy under clause 4. I understand that, but I am a bit concerned that the Bill appears to have no provisions to cap the charges for software updates. For the sake of argument, I have just spent £15,000 on a 2-year-old automated vehicle and then some software update comes in that is £1,000 and a month later they want another £1,000 out of me. If I do not do it, the thing is useless, because it is uninsurable and therefore undriveable. Do you think there should be provision for a cap on software update costs, so that vehicles do not become uninsurable and therefore driven without insurance?
I have not really got a view on a cap, per se, but I have got a view that if it is a fundamental safety upgrade and it will change the functionality of the vehicle significantly, there needs to be an arrangement in place to make sure that is not optional. It is probably for other stakeholders to say how we make that affordable to the public. From an insurance perspective, we do not want cars to be unsafe simply because people cannot afford safety upgrades. That is true today, thinking about automated braking: it would be great if that was a standard feature of all new cars because it is proven to be safe. It is optional and it is often not taken up because it is too costly.
Q I just want to return to two groups that miss out on the freedom and opportunities of being able to drive. We talked about older people and disabled people but also young drivers, for whom insurance is often prohibitively expensive, running into many thousands of pounds. What analysis have you done of the advantages of connected and autonomous vehicles over and above taxis, private hire vehicles, getting an Uber? What extra benefits do you see those two groups being able to derive once this technology is established and there is widespread take-up? Have you done any analysis or thinking on the social benefits for those two particular groups?
We have not done a research project on this, but I am aware that new products enabled by connected systems are opening up the ability to drive to a wider range of people. For example, younger people now have access to a wider range of insurance products enabled by telematics than was the case previously. Certainly, there is innovation within the industry that I am aware of, which is opening up options for accessing insurance to younger people as well as to some other groups as well.
The potential limitation is that we do not know when this completely automated technology is going to come to market. We are assuming about 2021, but we cannot be 100% certain. There is a quite a gap until then.
Telematics, which Iain mentioned, are not directly linked because it is a plug-in the insurer gives you that is not necessarily built directly into the car, but that is probably the first step towards an insurance policy tailored much more around tracking what you as an individual do, rather than broader risk factors.
Longer term, we are talking about cars that will take away the most common human errors and make the road safer. Increasingly, insurance is going to be tailored around the vehicle rather than how the individual behaves. Where you are talking about younger drivers particularly, their behaviour is going to become less of a factor. So you would not necessarily be thinking about age as a relevant risk factor when you look forward into the future. For older people and people who are vulnerable and do not have access to cars at the moment, this is transformational. We probably have not done any more work than any other witnesses on the evidence of that.
Q I have a couple of questions. The first one is for Ben. We have conflicting information about insurance cost. Insurance cost could be much higher because of the repair cost and the lack of people qualified to do the repairs, as well as the cost of the car itself, but there are also expectations that insurance costs will come down because there will be fewer accidents. Is it fair to say that at the moment there is just not enough information to do accurate modelling to understand what insurance figures are going to look like?
I think that some people have tried to do modelling, but there is that uncertainty between those two things. We don’t actually know what the cars are going to cost on the market, and that is obviously going to be a factor in the insurance premiums as well. That said, our members are really enthusiastic about the technology. I think they all recognise that it is the future of driving. We don’t know exactly when it is going to come to the road, but it is going to happen. I think they are going to be very keen to be involved in it from day one, and to therefore be offering competitive products that people will want. So there is a market incentive to say, “Don’t make this too expensive.”
In terms of really detailed modelling on the exact price, we do not know enough. On the technology side, a lot of that is developing now. We are going to get many more assisted cars. They might not be fully autonomous and self-driven, but that technology is the same kind of technology that will eventually lead to automated driving. We have already started work on resolving the questions around how good the repair network is going to be, so it is not just a question of waiting for automated driving and then it switching over.
Q Are there any special requirements that insurance companies would need for different testing, for example showing that people are capable of using the software or, with a semi-autonomous vehicle, when the right moment is to take action? We know that some people have trouble switching from a gear stick to manual, so this is another quantum leap. Are there any special requirements that insurance companies would like to see, going forward?
I do not think we would have any, other than what interested parties in the road safety world would want. I think we want drivers to be well informed about what they have to do. They have to know how this technology that they are taking on the road works and be confident about when they can and cannot use it. That is probably going to need to be part of the driver testing regime. It is a valid question to ask whether the driving test that you take at 17, which never changes again, is fit for purpose when technology will potentially be upgraded on a regular basis. That is worth further consideration. I am not sure that is for this Bill, but it is definitely something we will need to think about before the cars are commercially available.
Q I just wonder what input your organisations have in the testing trials that are ongoing. I know that there have been four trials in different parts of England, but I am thinking of the bigger issues. If we look at it from a Scottish perspective, we have rural roads, single-track roads and different weather conditions. There are connectivity issues, which my colleague touched on earlier. What plans are there to review the tests that are ongoing? How much more robust do the tests need to be and how is that going to be rolled out across the rest of the UK?
My team actually oversees the research programme that is paying for the tests you mention, the four city driverless car trials. It is really important when taking forward the competitions to have as open a process as possible. We work closely with Innovate UK, the Government’s innovation agency, to design competitions around challenges where we think it is likely that the UK is going to be able to pull through developments in the research base into products that are going to be usable and commercially viable. The initial set of tests were in London, Bristol, Milton Keynes and Coventry. We anticipate having future rounds of competitions that will be open to anyone in the UK to participate in if they want to form consortium bidding.
Q You know that the Bill attempts to strike a balance between, on the one hand, doing enough not to constrain future development—indeed, to facilitate it—and, on the other hand, trying to determine what the schedule describes as an “unknowable future”. Have we got that right, or should we have done more? I draw particular attention to the relationship between connection and automation and the issues of privacy and security of data. Should we do more now, or is it enough that we take powers to do things when we know more later?
It is a really important question. The advent of automated vehicle technology will in time require changes to different parts of our regulatory system. We have heard about some of those already today. The trick is to try to find ways of targeting the areas where we think action is necessary now in order to unblock barriers, or where we know technology is near to market. We need to make sure that we have the framework in place to enable the safe use of that technology.
To some extent it is a question that different people have different views on, but we certainly consulted last year with a range of different stakeholders on the areas where they thought action was necessary in order to ensure that the UK was doing the right things to set up a framework. The area in the Bill was the one that stakeholders highlighted as the one that was most important to act on first.
In time we will have to have further steps in the process of getting our regulatory framework ready. In doing so, I would hope to follow the same approach of identifying where the barriers are that need action now and which technologies are nearer to market. We need to make sure that we have the framework in place to enable those.
Q Can I go back to the definition? At the start of the session you said that the thing you welcomed in the Bill was that it would define what an automated vehicle is by whether or not that vehicle was on the list produced by the Secretary of State. Do you think that creating a definition will be simple? Where would autonomous emergency braking come into that? A large number of vehicles might have autonomous emergency braking that one would not normally define as automated vehicles. Nevertheless, autonomous emergency braking, by its nature, will take control of the car and stop it whatever the driver is doing. So would the car fitted with autonomous emergency braking need to appear on that list, because it would
“in at least some circumstances or situations” be capable of driving itself without having to be monitored by an individual? If it were included, are we saying that this new insurance product that the Bill brings into effect is essentially going to be the norm, not the exception, much more quickly than we thought?
Autonomous emergency braking is one of a suite of technologies sometimes referred to as advanced driver assistance systems. The Bill does not seek to set out a regime to manage those systems. It is about automated driving in vehicles where the driver can step out of the loop and does not need to be involved in monitoring the system. The difference between those systems and ADAS systems, as they are sometimes called, is that the driver always has to oversee what is going on in the vehicle. For those sorts of systems we anticipate the current regime being appropriate.
Q Is the boundary between those two as exact as you say? In a sense, with autonomous emergency braking, the driver has to monitor it. Whether the driver is monitoring it or not, that technology will take control of the vehicle.
We anticipate the measures in the Bill interacting with other aspects of law, including type approval requirements for vehicles, which will be looking at how different systems should be approved for safe use on the roads in this country. There is a lot of technical work to do to understand what the particular approval regimes will be for different forms of technology, but we anticipate the higher levels of automation that we are targeting in the Bill being different and distinct in the way they are approved from the ABS system that you were talking about.
I would anticipate quite a lot of work at international level to set the regulatory framework and technical standards that will underpin the safety framework for approving these vehicles. When that happens, there will be a decision for Ministers to take about how they consult with stakeholders in the UK to make sure that people are comfortable with those definitions before they are transferred into UK law.
May I ask a couple of questions relating to the way that you have looked at the insurance? It seems to me that you are treating the concept of ownership as it is today, rather than as it is likely to become; transport is likely to become a service, rather than a commodity. Is that fairQ ?
The policy aim of the Bill was to set up a framework that protected innocent victims of incidents relating to these vehicles in such a way that it felt similar to the current framework. We can have a framework around vehicle sale that is based on current patterns of ownership. In future that might change, as you say, in which case we would have to review the framework to make sure that we were making appropriate provision in law to allow people to operate the system safely.
Q It will probably become unlikely that car companies will end up selling their cars; they will lease them for shorter and shorter periods, as many car companies already do with their corporate fleets. It would seem sensible to have a look at that.
Perhaps we can go straight on to insurance. The safety systems before full autonomy—what you are calling level 4 cars—
Q Various cars, while not fully automated, already warn you if you are going to cross a white line or are getting too close to the car in front. As automation levels come up, are insurance companies intending to offer better premiums?
I would say that insurers have already done that. Autonomous emergency braking was referred to. Even before we had any claims data to back this up, we set any car that had that technology a lower group rating. If you have that technology in your car as standard, you get a cheaper insurance premium. We now have evidence to back that up; we have pretty robust data that say that that technology works. That is definitely the intention, going forward.
One of the key things that we as an industry need to know is when that technology is in a car. That is a practical challenge that we have. I do not think it will be a problem in four years’ time, when the Bill comes in. We as an industry would really like to know when this technology is in cars, to make sure that we are pricing accuracy. It is a data-sharing challenge, because it is often impossible to find out whether we have got it.
Q As the Bill comes in and starts to make greater provision for understanding who is liable, the question of ownership kicks in: is the driver responsible for upgrading the software, or does Toyota or whoever maintain ownership throughout? As semi-autonomy moves more towards full autonomy, you get an opportunity, but you also get this question: at what point do you start pricing out real drivers of real cars, if you see what I mean?
You do, potentially, but bear in mind that there will be a tipping point at which there are so many really safe cars on the road that it will have an overall impact on the number of accidents. The number of accidents will go down across the board. Also, the whole fleet will get safer; there will be a decreasing number of people in cars with no automated function at all, and even they will get the benefits of generally safer roads.
Q Of course, as the number of incidents goes down, premiums will presumably fall for everybody. Given that car insurance is the most lucrative area of the insurance market, have you done any work on what this will mean for house insurance and various other forms of insurance, on the grounds that it seems unlikely that your members will voluntarily lay profit aside?
Q Returning to the issue of software, clause 4 devotes a lot of attention to when insurers will not be picking up the can—something that we are familiar with. Can you say a little bit about how you are expecting software to be updated? What is the process for doing that? We all update our phones; we plug them in and press “install”, and the phone tells us when it is done. What is the current state of knowledge? Where are we, scientifically, on achieving that?
Linked to that, what responsibilities should there be on manufacturers to provide updates and tell the owners or users of vehicles that those updates have to be made? As I read it, there is nothing in the Bill that places any obligations on manufacturers to do that. A lot of time is devoted to when the software has not been updated, but where is the principal obligation for the manufacturer to do it? There are a lot of questions, but I am wondering whether that loops back to the definition and whether that needs attention to ensure that we have addressed the obligation. So how is it done and what are the obligations on the manufacturer?
Those are good questions. To answer the second one first, what is important about this Bill is that it is looking just at the insurance regime for these vehicles. It will have to work in concert with other parts of the law, including the system by which vehicles are approved for sale. You might imagine that if vehicles that operated automated systems were to be approved for sale there would be a close look at what would be necessary to ensure that the systems were updated where necessary to take account of any changes that were important to ensure safety.
Q Have we got this the wrong way around, then? Surely you have to establish how something happens before you start dealing with its insurance consequences. This is putting the cart before the horse, isn’t it?
We are focusing on this now because this is an area where in consultation people told us that it was important to set out a framework now to allow insurers and manufacturers to have those discussions about what might be necessary to inform the products that come to market when these vehicles do in four or five years’ time. In the meantime, we need to be working very hard to ensure that the appropriate approval regime for these vehicles is also in place. The vehicles will not come to market without that, so this will have to work in concert with another part of the law, which will say how these vehicles will be approved for sale.
If I can add one other thing, I think that the Bill is intended to do a new thing by protecting someone who is in the driving seat as, because they are not in control of the vehicle at the time of the incident, they are being treated as a victim. If they have done something to the car that means that they are responsible for the accident—perhaps they have not maintained it properly—it is reasonable to put it into their insurance policy that that is not something that they could claim for, as they would not be a victim. That is what these policies are broadly intended to do. I take your point that we absolutely need to define what updates need to be made and who is responsible for them, but if you turn it the other way by insuring the person in the driving seat and ensuring that they can claim if they are injured, the situation changes if they caused their own injuries.
Q If we are getting into a discussion on clause 4 about failures to update software, where is your starting point? You are basically telling me, “Ah, we’ll do that somewhere else separately. We will have to get those regulations on board.” All that I am suggesting is that that is out of sync and we should be looking at the processes first, at least for what we are expecting, before we start dealing with the insurance consequences.
To answer the first part of your question about how this is done, that is likely to develop over time as new systems come to market. It is already the case that some manufacturers upgrade software systems by asking customers to take their vehicles to a dealer and some do it over the air, in a similar way to how a phone is updated, for example. That is an area that is currently the subject of international discussions, and indeed the UK is co-chairing the international regulatory group that is having a look at how over-the-air updates will function in future.
Q Really, what the manufacturer would say is that if a vehicle has not been updated for one reason or another, or if they discover some other technical reason to shut it down, they will make sure that vehicle does not shift. Is it within the contemplation of the industry to take it that far?
Q I have a couple of questions. We have a problem in this country with uninsured cars. Given that these cars are likely to be connected to the matrix in some way, do you think that it would be sensible for the Government to take a power to require that the car has to check whether it is insured before it moves? When I go and buy my tax disc, the system checks that the car is insured before it allows me to do so. Should these cars be required to do the same?
Q You seem to imply, Mr Howarth, that the insurance industry would be indifferent to the characteristics of the passenger, who need not be the driver of a level 5 car.
I am a big fan of the rural pub, but I do not know the answer for certain. That is probably also an infrastructure question: I can see the cars working in certain inner-city areas, but personally, I am not 100% sure whether level 5 is ready for some rural roads yet. I think evangelists for level 5 technology will say that it is.
I do not know. I wondered whether you guys discussed it because the automated vehicle elements of the Bill do not apply to Northern Ireland, yet one would expect people in Northern Ireland, as elsewhere in the United Kingdom, to wish to have automated vehicles available to use.
Q I assume that you looked at other countries as you prepared for the Bill. Will you say a little bit about how other countries are addressing the insurance and regulatory challenges?
The legal frameworks in different countries are often specific to those countries, so it is not possible to do an exact read-across, but we are looking at what people are doing to see whether there are broad lessons that we can learn. For example, in California, if you want to test automated vehicles, you have to put up a surety bond to ensure that there is a provision to cope with any accidents. Looking at that and other systems, we felt that the system in the Bill was appropriate for the UK and how our insurance system operates. It builds on a system that people would recognise, so it would look similar to what people do now, and it targets an important policy, which is to ensure that innocent victims caught up in an incident involving a vehicle in automated mode can get quick access to claims.
I was going to mention European partners, but from an insurance industry perspective, I think that we are ahead of everyone else in having clarity about how the legislation will work. Obviously there are still things that need to be done before the technology goes to market, but I get the sense that other people are debating the issues, but not with a formal proposal on the table. I genuinely think that we are a step ahead of everybody else.
Q The reason this matters is that a lot people will now be thinking about booking their ferry and Eurotunnel tickets. Will we be able to take those cars abroad?
Which is part of the reason why it is important for some of the discussions about the regulatory framework to take place at international level, under the United Nations Economic Commission for Europe or other bodies that regulate how vehicles operate to ensure that, where possible, we have interoperable systems.
If you are thinking about cross-border insurance, as long as the broad principles are united—there are already big differences between the UK and other parts of Europe and how they insure vehicles; we have a driver-centric version whereas a lot of other European countries have a vehicle-centric system and a form of strict liability with various definitions—one would hope that we could evolve a system that gives at least minimum cover on a unified basis. We should not therefore have too much of a problem.
Q Mr Tugendhat made an interesting point. It had not occurred to me, but if I am in my automated vehicle, which I have taken through Eurotunnel, and I am driving down a road in France and a non-automated vehicle is coming at me in the middle of the road, I do not want my British automated vehicle diving off to the left—which is what you would do to take evasive action in this country—
This is a serious point in the context of Mr Forbes’s discussing interoperability. I presume there has been a discussion about the coding—I would like reassurance about this—so that the evasive action that automated vehicles might take when faced by unsafe manoeuvres by non-automated vehicles is appropriate to the side of the road on which one drives. Otherwise, we will have a big problem, as Mr Baker will know, with software coding and so on.
These are the sorts of challenges that you have to work through when you sit down to think about how the system will operate in practice. We are still at the stage of the technology where the developers are making sure that they can get their systems to work in particular locations—particular cities or areas. If the developers want to sell products and services that can be used in more than one country, that is something we will have to bear in mind when taking forward our development programmes. Indeed, if they are going to operate in accordance with the right regulatory framework, they will have to have discussions with regulators about how that will operate in practice.