April 30, 2018 | Written by: Patrick Sheridan
Categorized: AI | Claims | Industry Insights | InsurTech
Share this post:
In the last year, we’ve seen a number automobile manufacturers take big steps towards releasing widely available, fully autonomous vehicles. This has led to speculation that we will see a massive decrease in the number of automobile accidents, and this also has led to the big question in the industry, do we need traditional auto insurance if cars are no longer involved in accidents? Obviously, this is a key discussion around the world for insurers because the immediate next question is, if we don’t need auto insurance, what happens to the auto insurance companies? This question has significant economic importance; auto insurance is a $200 billion global market.
Insuring who, or what?
This is soon to move from a theoretical debate to real questions for the industry. These questions include who do we insure for an autonomous vehicle? Is it the driver, the car manufacturer or the software writer(s)? This is only the beginning of the list – undoubtedly the industry will change, and a safe bet would be that the auto insurance companies are here to stay for the foreseeable future. Further, them growing in the short term is just as likely. Breaking the discussion down to smaller parts, here’s why.
First, let’s put the premise on the table that the self-driving cars will be smart enough not to drive into one another. While that is a safe premise for marketability, it is not exactly reality. Volvo has announced that it will insure its self-driving cars, and that Volvo itself will cover the costs of any car repairs associated with an accident when in autonomous driving mode. While these cars will be introduced in various modes over the next few years, the number of models will be small for a while. Tesla is leading the industry currently, they offer a self-driving mode on all of their cars and trucks. Cadillac is advertising a self-driving mode for the CT6 as part of the 2018 model offerings, Volvo is claiming 2020. Many other automakers are working on it, but are in various stages of design, testing and implementation.
How long until we’re completely autonomous?
Understanding those broad timelines, will autonomous driving vehicles replace todays cars in the near future? The simple answer is not with current trends. Let’s use the US market as a sample. Wikipedia indicates that there were over 263 million passenger vehicles in the United States in 2015. This would exclude many trucks, motorcycles, and other vehicles used for transportation. There are approximately 17 million cars sold in the US each year. Using these raw numbers, current replacement rates indicate that autonomous cars would not replace current vehicles for 15 years. Simple observation shows us that there are still a reasonable number of cars manufactured in the 1960’s and 70’s on the road today. This makes any complete take over by autonomous vehicles likely to be more than 50 years away.
What happens when there is an “accident?”
Next, let’s look at accidents. While the premise that autonomous vehicles will not run into one another could have merit, recent news proves that “no accidents” are unlikely. Just recently, an autonomous vehicle hit and killed a pedestrian in Tempe, Arizona. There are arguments on both sides as to whether the technology should have detected the pedestrian and stopped the vehicle. Regardless of fault, this tragic accident will lead to improvements in the technology, but as the situation shows, “accidents” are still a reality.
Let’s also discuss what makes up auto insurance. Typical auto insurance products include coverage for comprehensive damages for things the owner or driver can’t control. For example, hail damage to the body, a tree branch falling on a car and causing damage, someone running into the car while parked, theft, vandalism, etc. While there are always improvements to minimize these damages, these risks can’t be fully eliminated. Unless autonomous cars are going to be significantly cheaper, your average consumer will still want protection from these perils.
Paying careful attention to auto manufacturers press releases about autonomous vehicles they all refer to “autonomous” as a feature or mode for the car, which implies there are other modes, for example, being controlled by a human. Therefore, there’s still a very good chance that if there is an option for humans to control a vehicle, there will be accidents. So, barring any major change in our legal system, as long as there is the risk of an accident, the mandate for insurance remains.
Another common notion is that with more self-driving cars on the road, there will be fewer accidents. Fewer accidents imply fewer claims. But can we let the logical argument progress that fewer claims equate to lower repair costs, leading to lower insurance premiums? Perhaps, but it is not very likely. There has been nothing specific that indicates these newer autonomous vehicles are cheaper to build or maintain than current vehicles. All the technology required for autonomous driving may be more expensive to fix than today’s standard vehicle. There is also the unknown exposure of the legal system. Who is responsible if a vehicle moving in autonomous mode is in an accident? Is it the manufacturer? Is it the autonomous software developer? Is it an independent testing service? These costs most likely will have to be covered by insurance companies and indirectly lead to higher premiums.
While these simple logic arguments imply that the need for automobile insurance will not go away any time soon, it doesn’t mean the industry won’t change. Certainly, as the number of autonomous vehicles increases the number of accidents will decrease or at least the severity of accidents will decrease. This trend will also be impacted by things like urbanization, shared cars and remote workers. The combination of these factors implies that that the size of the traditional automobile insurance market will decrease—but not go away. Open for discussion is will this lead to new types of insurance designed for the unique circumstances surrounding autonomous vehicles?
Industry must embrace change
A summary of this logic implies that the industry is likely to consolidate. There will be a smaller number of people paying premiums, fewer severe auto-related accident claims, potentially higher repair costs of more complex technology, and a larger number of “products” to address the combination of coverages that the public will want.
An assessment indicates that an insurance company today, planning to be in the auto insurance market in 20 years need to be making the following adjustments now to streamline operations to be more efficient, specifically:
- Modernize core insurance platforms to allow for easy introduction of new products (groups of coverages)
- Move the appropriate pieces of technology platform to the cloud to take advantage of economy of scale when growing or shrinking
- Enhance claims processing to allow for efficient processing of less severe non-bodily injury claims, streamlined further by using cognitive technologies
- Build a solid base of policy holders’ profiles today so as market trends change, the insurance company can model the impact on its specific book of business and react to changing market conditions faster than competitors.
While the coming explosion of autonomous cars will not eliminate the auto insurer we know today, it will certainly force change. Given the volumes that insurance companies deal with and the historical lack of success with large projects, time is of the essence to prepare for the changing market now.
Learn more about IBM Insurance solutions here.
Have any thoughts on how self-driving, autonomous automobiles are and will change the auto insurance landscape? Let us know in the comments.