Hi, everybody. Today, we are joined for this interview by Daniel Vignon. Daniel is currently a doctoral candidate here at the University of Michigan in the Department of Civil and Environmental Engineering. Daniel will be joining the faculty at NYU in the fall. He is an expert in economic modeling and issues around data and privacy that arise with respect to connected and autonomous vehicles. We're delighted to have him today. Daniel, welcome. Thank you very much, Liz. It's my pleasure to be here. Let's start talking about risk and safety. Connected and autonomous vehicles are expected to greatly reduce the risks of deriving. Should we expect the reduction in risk to be passed on in terms of savings in insurance cost or other savings to drivers or to society? That's a very good question. I would say not necessarily. For example, for a few years now, we've had ADA system, so advanced driving assistance systems like blind spot monitoring, lane keeping assist. Even though they've been out for a few years and there's research that shows that they can improve the safety and reduce the occurrence and severity of some crashes, it has not translated into lower insurance cost for owners of these vehicles. In general, that's because it takes a lot of time for the insurance industry, for example, to evaluate really these benefits of the systems. It takes a lot of times for the system to penetrate into the market. It takes a lot of time to assess and to disentangle the different effects and factors that are orienting to a given accident occurring. Already we see here that better safety or better instrument doesn't necessarily translate into lower cost. Another thing that is important to consider is also that we've automated connected vehicles. You might have actually higher repair costs. That needs also to be taken into account. You have more complex systems and naturally it takes more time, more effort to repair them when they fail. Taking all that into account, the result of the introduction of connected autonomous vehicles might not be actually lower cost. But I think it's also important to talk about some of the positive that might come with that. We've seen, for example, Tesla recently rolled out its own insurance program. What they've been doing is they're able to monitor drivers and they say, well, we know our systems better. We can monitor drivers, and we expect that, for example, an average driver might be able to save 20-40 percent on their premiums if they were to contract the insurance with us. There is that opportunity here for cost reduction that might be possible. I would say the last thing, I think, when we talk about autonomous vehicles, we might instead consider talking about cost of ownership rather than cost of insurance. Because since the liability here of who is responsible for an accident might shift. It might not so much be the driver anymore but more so the company, and in that context the company would have to contract its own insurance and might then transfer that into the cost of the vehicle. So from a consumer perspective, what I'll be seeing is the cost of owning that vehicle overall. This might, because of all we talked about, go up or go down. So interesting. Certainly some factors that may be increasing cost, others that may be reducing cost, not likely really to see any net changes, at least in the near term and it'll be interesting to see how that plays out over time if there are reductions that are able to be passed on to consumers. But it sounds like it's anybody's guess whether it goes up or down at the current. Excellent. Thank you. You mentioned data that there's data that may help on the one hand insurance companies better assess risk and so on that's associated with connected and autonomous vehicles. Of course, data is a big deal in all of this discussion. Who owns all that data? A lot of data is being generated but who owns it and what does it matter in terms of who does own the data? Here again, a very good question. I think when it comes to what the law says both on the federal and state level, there is no real uniform regulation or way in which data is looked at. So this give rise to some interesting things. I think Mark Zuckerberg when he was testifying in front of congress last year, he said, you own your data, which I'm sure is true. You maybe sign away that right or give authorization to Facebook. But often you really don't have any control over what happens to that data, once it's collected. But some companies are starting to try to use that as a marketable item. Turning up with Apple is saying, as commercials now saying like we respect your privacy, we don't share your data. I think Tesla for its ominous autopilots monitoring program asks you to voluntarily signing. This is one dimension where there hasn't been any regulation. Companies do what they want. Some tend to want to give you more power as far as how you handle your data and some might not. But I think another aspect also that we might look at is say, Tesla collects that data, or say an insurance company. See, progressive collects data on your driving because it's only device in your car. That surely gives Tesla all progressives. It's an advantage in terms of pricing and offering new entrance because they actually can evaluate your risk that other companies might not be able to. Here, one of the question we might ask is, well, should they be forced to share that data with their competitors? Again, make it possible for you to get even lower rates because now they all have the same information they can compete for you. There is a research that suggests that doing that, for example, might actually discourage a company from doing that. Because see, when progressive decides to try to enroll new in the monitoring trunk program, they have to give you a high discount at first to incentivize you to join the program. There is a cost associated with that. It's not free. By forcing them to share that data, you might be in a position where they really don't have any more incentive to do that collection. From that perspective, who owns the data? What they can do it also matters, and something that definitely needs to be looked at. Yeah. It seems like there's a lot of opportunity for policy that would create incentives or regulations that would require both companies and maybe individuals to share data in different ways. Are there policy approaches out there that you're aware of or that are being developed that you think would be beneficial in terms of sharing the benefits widely of the data that will be and is already being collected, and generated through autonomous vehicles? Actually, I think that there is much that can be done. I think when looking at it across the spectrum of policies, I think that there are two extremes. On one end, there are policies that might give you the power to decide. For every single use a company wants to make all my data, I have to get approval. That's one extreme. Another one would be a system where you really have no defined rights of what might happen. I think both is what we want to avoid because, for example, if you have to give approval for every single use, I don't think necessarily we have the bandwidth to do that on one hand. It really introduces a lot of friction in the way in which you can do any information because it's important to have information flowing freely. It benefits you and it benefits other people too. But of course, we don't want to be also have to add the other hand where you really have no sense of what's happening with your data and what is out there. I think one of the first things that can be done at this level, I think both at the federal and state level is to really have a better definition of what companies can do with your data. What they need to notify you for? When do they need to notify you? Hey, this is what we're doing, the data right there. There's going to be certainly important aspects that needs to be done. Also, even in terms of awareness, are you aware? Oftentimes we're not even aware. Like this is what we're given now. I think it's defining those parameters. I think earlier on would be important. As far as wonderfully distress, and I think in Europe they have the GDPR. I've used the GDPR so much that I forget what it stands for. We can look that up afterwards, but they have some law that gives you a bit of control over certain aspects of the data to them. That would be an example of regulation. Maybe a model that we might look to in the US or even other countries as an example. Exactly. Yes. Daniel, thank you so much for sharing these insights, wonderful ideas to think about. We really appreciate you joining us here today. Thank you. Thank you very much. It's my pleasure.
