Future intelligent transport systems and risk handling in the insurance industry: Are we ready for autonomous cars on European roads? 

Av Hilde Waaler

The development of autonomous cars is rapidly progressing. As is the development of cars with increasingly high levels of computer technology integrated within the vehicle. How can we handle the risks of these developments? Who is responsible when incidents happen and what will be the role of the insurance industry in future mobility? We have talked to Reto Schneider, former Head of Emerging Risk Management and with 20 years of experience from the Swiss Reinsurance company Swiss Re, about some of the challenges more intelligent and autonomous transport introduces. Schneider assists the SIITS-project (Vulnerabilities in future integrated intelligent transport systems) with competence regarding autonomous mobility solutions and the role of the insurance industry.

One of the aims of autonomous cars is to drive safer with reduced risks of incidents and injuries. An increasing degree of computer systems integrated into the vehicle’s driving capacities will, according to EU, reduce car accidents and make driving safer. However, more complicated technology and more computer-related processes also creates several new issues. We see challenges related to who has the ownership of the car data, undefined areas of responsibility as well as challenges regarding standards and regulations. New technology also allows for an increasing degree of digital attacks and cyber risks.

“We see several challenges and increased complexity related to the development of autonomous vehicles and risk handling,” says Reto Schneider.

Who owns the data generated by cars?

“One of the challenges is who owns the data,” says Schneider. “Is it the OEM (original equipment manufacturers) or the software suppliers? The drivers or the car owners? Those who have access to data are not always willing to share this information with other companies. For the insurance industry this creates challenges,” he underlines. “It creates an asymmetry with regard to the information the different actors have access to. This is a problem for insurance companies because it makes fair risk-based pricing difficult.”

Some companies change their business model in order to accommodate this new development. For example, the car manufacturer Volvo recently purchased the software company Zenseact which develops computer programs for autonomous driving.

Several new actors – undefined interfaces

Many new actors have recently become involved in the mobility industry, new OEMs, suppliers, software producers, etc. In addition, new insurance solutions based on new business models have been introduced. Undefined interfaces between the different actors create increased risk and undefined areas of responsibility. Many of the new actors are not subject to neither Norwegian nor European law. This complicates such interfaces even more with regards to responsibility.

Many companies are going to offer total mobility packages, including insurance. These companies may be competitors to the existing insurance industry or become a new client category.

insurance picture

“How does this development affect the insurance industry?”

“Currently this development has few consequences,” says Reto Schneider, explaining: “Car insurance as of today follows a simple principle. The owner of the vehicle is the responsible party, independent of who is driving the car. Strict liability still applies.”

“However, this is changing,” underlines Schneider. “The car manufacturers aspire to have more control of the entire value chain. In the future the car producers will not sell cars to individuals. They will sell mobility. Many car manufacturers are planning to provide large fleets of cars with integrated insurance.”

“With these large fleets, the risk will accumulate,” underlines Schneider. “And the number of policies issued will largely be reduced and the premium volume will drastically shrink.”

“And in case of an incident (Software failure, hacker attack, lost GPS signal, etc.) several thousands, maybe more than hundred thousand cars may be grounded simultaneously. However, I do not expect this as an immediate threat. The transition period will probably last for several decades.”

This development also creates uncertainty as to who is responsible for acquiring insurance. Because the driver is no longer considered to be the “driver” but is more often defined as the “passenger”.

“With the shift to product liability, who will be responsible for driving the car safely? Will the OEM take responsibility for safe driving as well?” 

“This will depend upon the automatization level. With high levels of automation, the driving responsibility can no longer be with the “passenger”. As long as human drivers are expected or able to take over the controls, the responsibility will remain with the human driver. Currently the law applies strict liability, and the owner of the car is liable in the first line, but if another person is “guilty”, the insurance company might initiate a regress.”

Five levels of autonomous driving

“How does technological development influence the insurance industry?” 

“The Society of Automotive Engineers International (SAE) has developed an internationally recognized overview of automation levels, which includes five levels of autonomy in vehicles:

  1. No automation
  2. Driver assistance
  3. Partial automation
  4. Conditional automation
  5. High automation
  6. Full automation

Reto Scheneider fra Swiss Re
Reto Scheneider

“Driver assistance and partial automation have been around for a while. Focus is now on level 3. Mercedes-Benz and Audi are probably the OEMs that have obtained the highest development levels. However, the development is proceeding slowly”, says Schneider. “Level 4 and 5 will probably not be introduced before 2040.”

“Are we ready to roll out the autonomous cars onto the European roads today?”

Not at all. The cars are not ready yet! I would be happy if traffic lights could be controlled with AI and allow for dynamic management of traffic flows”, says Schneider.

“What is the biggest challenge?” 

“The biggest challenge with regard to risk and insurance is related to level 3”, says Schneider.” Mixed traffic where the driver must be prepared to execute his driving skills within seconds will create increased risk for incidents,” he says.

Ego cars

“Another concept launched is the Ego car,” says Schneider and explains: “The Ego car is an automated car, fully autonomous, fully packed with sensor, computers and communication devices allowing Car to Car Communication. An Ego car do not depend on public infrastructure to steer and control and optimize the driving patterns. In contrast, a public transportation shuttle fully controlled by a public infrastructure without a driver, would represent the other pole.”

 “Who will be liable if an autonomous vehicle drives into a group of people”?

“For the insurance industry this will make little difference from today’s practise. As of today, the owner is responsible. In the future, if the car is driving the car, the car will be liable. And liability will shift from today’s strict liability to product liability. We do not have to change our laws to be able to handle automated driving. One problem remains though. Under strict liability the compensation of the accident victims might be much faster compared to “claims payments” under product liability regimes. This is an area that must be redesigned to assure that the money flows quickly to the injured party.”

High-tech vehicles

According to Schneider, the cars of the future will be packed with computer technology and high-tech sensors. The degree of autonomy will increase, the cars will be safer, and we will see fewer ”small” incidents on the roads. However, such development also has a potential for larger incidents affecting many vehicles. Not least can we expect that the risk of cyber incidents will increase. These are all aspects that will influence insurance premiums.

“Besides, computer technology is expensive,” says Schneider. “Even the smallest accident will cause damage to several sensors. Repairing will consequently be expensive. This will affect the cost of the insurance and increase payments after accidents. We might see more “total losses” because the repair of the sensor technology is more expensive than the residual value of the car.”

New smarter insurance solutions 

In parallel with technology development, the insurance industry develops new solutions in order to accommodate new demands and a changing risk reality.

“New business models drive the demand for new insurance products,” says Schneider. “Several insurance companies are exploring new possibilities for smarter insurance solutions based on big data. One example is the User Based Insurance (UBI) also known as Pay-as-you-drive (PAYD) or Pay-how-you-drive (PHYD). This is insurance based on driving length, car model, driving attitude, etc.,” says Schneider. “Other solutions include the sharing of cars between different drivers and new concepts like the ‘neighbour car’.”

Digital attacks and cyber insurance

New technology provides new opportunities and safer cars. However, with more digitalization the possibility for digital attacks against both the car manufacturers and the fleet owners creates a greater potential for damages. New technology allows for new methods of attacks. Cyber security will consequently be an area that OEMs, software producers and the insurance industry will need to relate to in future transport solutions.

“Who will the drivers address if cyber incidents occur? Who will be responsible in case of cyber-attacks?”

“This is not an easy question to answer,” says Schneider. “In fully automated cars, level 4 and 5, I assume it is a product liability issue for the OEM. It might be a different situation if the public infrastructure is attacked, and certain services will be out of operation. I fear that the recurring problems or losses will not be insured, and the burden of these losses will have to be carried by the users.”

The infrastructure is not ready

There are many issues that has to be addressed before autonomous cars can be released on public roads. One such challenge is the dependency of satellite signals in autonomous cars. Global Navigation Satellite Systems (GNSS) are used as a source for positioning and navigation.

“Today almost all modern cars have installed GNSS-receivers”, underlines Schneider. “Autonomous cars will totally depend on positioning data in order to navigate.”

“What happens if the satellite signal disappears? Who is responsible for incidents caused by signal loss?”

“The owner of the car is still responsible here. Whether a company or a person owns the car. If the signals disappear, one needs available back-up systems, like manual maps, other types of aiding systems, etc.”

However, to be able to fully exploit the advantages of digitalization, the infrastructure must remain fully functional. The protection of traffic management infrastructure and sensor technology in cars from malicious attacks is a new field, and security solutions have to be developed.

“What will happen if the infrastructure itself is responsible for an incident, asks Schneider.” “Who will then be responsible? Who owns the infrastructure? The society? Will we buy mobility services from our city or community we live in?”

The role of the authorities 

“What can we expect from the authorities?”

“The insurance industry awaits regulations from the authorities. However, there are many challenges related to standards: different types of vehicles, different global car manufacturers, infrastructure, etc.”

“New technology requires cooperation across borders. The authorities need to cooperate, and standards which apply across borders need to be developed. Cars driving on European roads need to relate to European standards and be able to cross the borders without legal obstacles or other societal obstacles.”

“In addition, the sharing of data needs to be accommodated in order to make data more accessible. Many countries have laws that protect data. But we will need laws that allow for increased use of data, whilst protecting privacy issues at the same time. There are still unclarities related to the distribution of driving information, location data, etc.”

“The discussions on these topics have started, and the EU has launched several initiatives for sharing and providing access to data. But the harmonization process needs time” says Schneider.

“There are large uncertainties and risks related to the development of new mobility solutions. The degree of uncertainty as of today will probably need to be reflected in a higher premium / uncertainty charge,” he says.

“In the continuation, the insurance industry will assist the development of new mobility solutions. While the insurance industry is an innovation enabler, they are not an innovator as such. The innovation is coming from the car software and supply parts industry. The insurance industry has a major part to play as risk reliever. However, mobility as a service force us to rethink and rebuild commercial insurance,” concludes Reto Schneider.