How Vehicle Insurance and Autonomy Intertwined

In early 2023 Oxbotica claimed at an event, which was held at Lloyd’s of London about the Future of Autonomy that insurance and autonomy are intertwined.

At the event, Sam Tiltman, sharing economy and mobility leader for the UK & Ireland at Marsh, claimed that the combined impact of Mobility-as-a-Service (MaaS), electric vehicles and automation is huge. Their interconnectedness could disrupt the insurance industry if autonomous vehicles were to prove their worth by delivering on their premise by preventing accidents, injuries and by reducing insurance risk. To avoid or to prepare for any disruption, Tiltman advises insurers to start investing now.

Adding to this is the prospect of changing from having one vehicle operator or driver to have zero vehicle occupancy. Autonomy means that the traditional paradigm of how a vehicle is controlled is going to change, making it possible to have ‘n’ operators per vehicle. Insurers will need to calculate how many operators amount to ‘n’ as they would need to decide what it means.

Insurance is an enabler

Rebecca Marsden vice-president of risk and insurance at Oxbotica says that insurance is an enabler for self-driving technology at scale. She believes it is mission critical to “unlocking the benefits of autonomy to every person and organization on the planet”. She also thinks that insurance is about underpinning consumer trust as it is a route to market for autonomous vehicles – adding that there is a powerful relationship between autonomy and insurance. In her view this isn’t just about risk management.

She explains: “The depth of usable data and extent of analytics available through autonomous vehicle technology is transformative.” Artificial intelligence (AI) can be deployed to validate and verify autonomous vehicle software for virtual world simulation, automated discovery of challenging scenarios, and real-time data expansion. The aim, she says, is to answer questions such as:

• What makes the AV safe?
• What external factors and challenging circumstances must it navigate?
• How can we plan for conditions that have never been encountered or envisioned?
• How can we build a validation and verification system that allows for rapid updates to ensure our product utilizes the latest technology and software?

Analog and digital worlds

To this, Julian Broadbent, CEO of Applied EV responds: “The insurance world is data rich, however; it spends most of its time extracting data from an analogue world. Connected vehicles are at the very edge of digitalization, they are all about providing data. This may be communicating legacy data or storing data. With connected and autonomous vehicles, the data is on a platter as they are so well equipped. As every step of the vehicle is digitized, they are able to produce amounts of data insurance companies can only dream of.”

However, you would be forgiven for thinking that all is rosy. Chris Moore, head of Apollo ibott 1971 reports that there is still some resistance to autonomy. It still suffers from having an unfavorable public perception. He believes that society accepts elevated levels of autonomation in many industries. Yet, autonomy incites fear whenever it comes to the prospect of increasing it on our roads.

Echoing Marsden, he adds: “Insurance plays a vital role here to improve this perception and create trust. Companies looking to deploy and operate autonomous vehicles are investing time with key insures in the space to develop fit for purpose insurance products that can enable this future of transportation to grow and flourish. That insurance will likely take the form of a usage based and embedded product. This will completely disrupt the distribution of motor insurance as we see it today.”

Significant impact

Despite this, he concurs that the impact of MaaS, electric vehicles and automation will be significant, citing the environmental impact of reducing pollution from the vehicles, although some critics dispute the claim that electric vehicles are as clean as they are said to be. Nevertheless, the transition away from internal combustion engines will, in his view, inevitably have an impact on society too.

He explains why: “Mobility-as-a-Service today via rideshare, micro-mobility and new leasing and car sharing services are already causing levels of disruption to vehicle ownership. Autonomous vehicles will be the catalyst for the vast majority of society giving up on vehicle ownership all together! As a result of that the distribution of insurance for transportation, and in particular motor, will shift from a personal lines purchase to a large scale and complex commercial purchase.”

Underwriters must adapt

Strategy consultant and Senior Manager at Ptolemus Consulting group, Alberto Lodieu, urges underwriters to switch their risk assessment methods away from analysing a human driver’s behavior to the autonomous systems own comportment. It’s no use resting on one’s laurels and relying on the claims that autonomous vehicles will be much safer than human-driven ones. They might be, but they still have a Level 4 of autonomy risk associated with them.

This risk will vary, he says, depending on multiple factors. They include the way automakers built the system in the first place, the environment in they are operating, the number of miles or kilometers the AV have been driving, the last time a vehicle had a systems update, the high definition map provider, the manufacturer of the different sensors included in the vehicle, the number of redundancy systems and numerous others.

He adds: “Analysing the performance of autonomous systems is not new. Insurers have been doing so since the emergence of ADAS (Automated Driver Assistance Systems). Today, most new vehicles have multiple ADAS, each impacting the car’s and driver’s risks differently. Thanks to pre-launch testing and relatively low adoption, insurers have incorporated the risk assessment of these systems into their scores. However, as vehicles become autonomous, the number of variables underwriters must consider significantly increases. Moreover, we are moving into a coexistence phase between the human and the autonomous system. To cope with these challenges, underwriters are moving from models based on historical data to predictive models using historical and real-time connected vehicle data.”

Propulsion models

That said, to what extent is it true to say, given that Germany wants to explore extending the life of ICE powered vehicles by exploring new and less polluting fuels, that the combined impact of MaaS, electric vehicles and automation will see the end of insurers policies for ICE vehicles? Well, Broadbent thinks that “the type of propulsion doesn’t affect any opportunities for vehicles to be autonomous, to bring safer and more efficient vehicles as well as methods of transportation of logistics, as well as the movement of people”.

He adds that there are a number of different solutions for clean propulsion. To this extent, while electric vehicles are often touted as the answer to reducing vehicle emissions, it is entirely plausible and possible that new clean versions of propulsion could either be commercialized or emerge “like what we’ve never seen before”, he says before commenting that with the search for alternative forms of propulsion, “there is a 90% solution, but the final 10% presents an extremely difficult problem to overcome”.

“For example, the use of hydrogen as a fuel is well understood, but hydrogen storage is logistically very difficult. At present, there is nothing that’s commercialized at scale. There is a way to store hydrogen safely to satisfy consumer needs: how to safely handle the fuel, the cost of the vehicle, and the ability to access real infrastructure. For that reason, we have seen early examples of hydrogen-powered vehicles but nothing that has satisfied all of the requirements.”

Data dependency

As autonomy is heavily reliant on data creation, storage and analysis, Broadbent sees insurance companies offering incentives to push everyone towards data-rich machines. This would certainly interest insurance companies as they would have a better insight into risk, which should translate into an opportunity to provide more affordable insurance. Meanwhile, Marsden concludes by citing data from the World Bank, which suggests that if countries don’t invest in road safety, they could miss out on “anywhere between 7% and 22% in potential per capita GDP growth over a 24-year period.”

However, while the intertwining of autonomy and insurance should reduce risks, which are often caused by human error in traditional vehicles, and lead to a reduction in incident severity as a result of increased response times, Moore says there could be an inverted reaction in some jurisdictions. This could lead to the severity of loss increasing. For example, in the US most injury cases are presided over by a jury, and so he believes that large autonomous vehicle companies will be “targets for reptilian litigation tactics and nuclear verdicts.”

Complex claims

This is because autonomous vehicle insurance claims are potentially very complex and therefore be hard to defend. This is why there are some insurers working proactively ahead of time to develop claims defence strategies. Complexity means that insurance, legislation, and autonomy must be intertwined. They rely heavily upon each other. Nevertheless, as Broadbent suggests, autonomy should reduce the number of vehicle-to-vehicle incidents and therefore reduce the cost of claims that insurance companies traditionally incur.