Shulou(Shulou.com)06/02 Report--

Produced by big data Digest

Source: bloomberg

Compilation: Liu Sijia, Xueqing, Aileen

If there were no more traffic accidents caused by human error, 90% of the risk would be eliminated.

The insurance company is busy preparing for this.

If you are considering buying a Tesla Model X, you need to pay an extra premium for this type of car with a self-driving system, in addition to where the charging posts will be located.

Dan Peate is a venture capitalist and entrepreneur in Southern California. When he was considering buying Tesla Model X a few years ago, he gave up the idea because of high car insurance.

"their offer was $10000 a year," Peate recalls, much more than regular car insurance.

This seems to be completely different from the driving concept promoted by Tesla and even self-driving: a series of sophisticated lidar, radar and camera opportunities are better at detecting dangerous situations than our mortal eyes and ears; and computers will never get drunk. click on Wechat or doze off while driving.

Self-driving wants us to believe that they will create a safer road environment.

However, concerns about recent accidents in various types of self-driving cars, including the troubles caused by some of Tesla's self-driving modes, make it hard to believe in the so-called advantages of self-driving cars. As a result, the price of relevant insurance has been raised too high.

For autopilot, the traditional actuarial model may not be applicable.

Peate says that when insurers and actuaries price a new type of risk, they charge more because they don't have enough data.

It is obviously more difficult to price Tesla's unmanned car.

Due to the lack of Model X on the road, there is a lack of safety record data.

A large amount of data is in the hands of Tesla and other automakers, which they hope to improve the degree of automation of cars.

On the other hand, insurance companies need to rely on the "accident data" obtained over the years to underwrite and set a true and accurate amount of insurance.

So why not get a lot of data from the industry so that you don't have to wait years to get "accident data"?

Deloitte foresaw this in its 2019 insurance outlook report. "the rise of connectivity. A large amount of real-time data has been generated and the relationship between insurers and policyholders has shifted from static and transactional to dynamic and interactive.

Tesla's Model S is equipped with autopilot.

For the auto insurance industry, which makes billions of dollars in profits, the shift means a bigger existential crisis. If there are no real drivers, why do we need car insurance?

Insurance premiums and company income are based on the likelihood of a driver's accident and the actual crash rate. More than 90% of accidents are caused by human error, and excluding drivers means a major change for insurance companies.

"this topic is discussed at every strategy meeting," says Michelle Krause, senior managing director of Accenture's insurance customer service division. Big insurers are "very focused on understanding the technology behind [automation] and what opportunities they will have."

Based on research from the Stevens Institute of Technology (Stevens Institute of Technology) in New Jersey, Krause's team released a report in 2017 predicting that insurance companies will face difficulties with the spread of automation.

Premiums are likely to fall by 12.5% in the overall market by 2035, although a new line of insurance products centered on self-driving cars will offset some of the losses, but the decline in premiums will eventually outweigh the gains, the report said. The good news for the industry is that it still has some time. According to Stevens Institute of Technology, there will be only 23 million self-driving cars on American roads by 2035-less than 10 per cent of the total number of cars in the United States today.

So far, the technical maintenance costs for self-driving cars have been very high, which means premiums will initially rise as more and more self-driving cars are taken offline from dealers. "considering so many sensors and calibration systems, a small collision can be costly," said David Ross Keith, an assistant professor of systems dynamics at MIT.

The auto insurance industry will no longer focus on consumers, but on companies.

As automation reaches levels 4 and 5-full autonomy, which can be taken over by a human driver or unattended-the insurance industry will change dramatically.

"it is foreseeable that the insurance industry will no longer be a mainly consumer-oriented industry in the future," Keith said.

That's because the driver himself is no longer part of the risk. "responsibility is likely to shift from individuals to software manufacturers and licensors that drive autonomous vehicles," says Rodney Parker, an associate professor of operations management at Indiana University.

Accenture's report agrees. This means that insurers will sell more policies to companies and fewer policies to drivers: drivers will no longer be compensated for not checking their blind spots, instead, carmakers and suppliers of communications systems, software and sensors will be affected by product failures.

More broadly, the nature of the risk itself will change, says Hyejin Youn, a professor at Northwestern University University. For human drivers, "uncertainty is random and random events obey normal distribution". If the risk lies in defective software or sensors, it becomes "more systematic".

National Mutual Insurance (Nationwide) has begun to think about this issue. Today's drivers are graded according to factors such as gender, age and driving experience. "if we get data from cars, the ratings will change dramatically and become very complex," said Scharn, vice president of product development at Teresa. Of course, insurance companies are already involved in data management, but "We have to show more power." Like the national mutual insurance company, Allstate is actively hiring big data and experts in the field of analysis, according to Don Civgin, president of its services business.

Insurance companies and car manufacturers will work closely together in the future

In this new world, when problems arise, it can be tricky to judge who is right and who is wrong.

If the lidar fails, is it the fault of the carmaker or the lidar supplier?

What if the driver didn't get the latest firmware update in time-so now it's his fault? If a Cadillac with a super cruise system loses its Internet connection, is the mistake made by General Motors, the carmaker, or Verizon, the telecoms company that provides network signals?

What if the car is hacked and transferred to the thief? What if the data on the municipal infrastructure that manages traffic flow is lost? "our society must figure out who is responsible for these different things, which will determine who should insure which risks," said Keith of the Massachusetts Institute of Technology.

Youn of Northwestern University sees it as a "public policy issue", "best solved by the government".

However, these are the quickest opportunities for traditional insurers and startups like Avinew to adapt. Keith said that policies to protect products will become more extensive, which also means that we also want to "ensure our own safety as passengers." If there is no driver, there will be no driver to be insured.

National Mutual believes that the smoothest way forward will be for insurers and automakers to share the steering wheel. "We are trying to build a deeper relationship with automakers," Scharn said.

Perhaps this means that a merger between insurance companies and carmakers is just around the corner? "the conversation is going on right now," said Krause of Accenture, a consultancy.

Related reports:

Https://www.bloomberg.com/news/articles/2019-02-19/autonomous-vehicles-may-one-day-kill-car-insurance-as-we-know-it

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.