The road towards the autonomous car and its insurance implications
Roland Voggenauer-Graf von Bothmer,
04 Feb 2015
Autonomous driving technology is already upon us. It is already starting to have an impact on accident frequency rates. Insurers will have to respond. How quickly they will have to depends on a number of factors, of which technology is only one.
Getting ready for change
As reported at the Swiss Re Centre for Global Dialogue Autonomous Car conference, held in September 2014, accident frequency rates across industrial economies are falling. There are many driving factors behind road safety. Better vehicle safety, with technology such as airbags and new braking systems, has certainly played a significant role. Improved emergency response and treatment have also helped reduce death rates.
That trend is set to continue. Current technologies, such as lane change, brake and parking assistance are the first steps in technology-enhanced human driving. Next generation technologies, such as automatic parking or highway cruise control, will see computers autonomously undertake parts of the journey themselves. All major global car manufacturers are working to develop these and similar capabilities.
Some industry players are taking a different approach. Most prominently led by Google, they are developing fully robotic cars that do away with the driver altogether. Still some way from commercial realisation, these fully autonomous cars are nonetheless clocking up test mileage and media interest. The tech companies behind these fully autonomous vehicles see technology adoption as potentially being like that of the smart phone; rapid and almost viral.
The ultimate vision of these fully autonomous car developers will be a whole new approach to our mobility, our lives and our societies. Proponents believe that the technology can be developed so that accident rates drop precipitously; indeed, that riding in an autonomous car becomes as safe as riding an elevator. Cars will be able to be summoned at the click of a smart phone. This will be conducive to shared ownership, fleets of remotely located cars, and the ability to select a particular car model dependent on the journey the passenger wishes to undertake. Congestion will be substantially improved and the need for parking greatly reduced.
Insuring the new market
The introduction of autonomous technology will have two notable effects on insurance. Firstly there will be a shift in liability. As cars undertake more and more driving functions, the driver will less responsible. One of the selling points of autonomous technology is that the driver should, at least for certain functions, be able to completely take their hands off the wheel and do something else. The second shift is data. Car underwriting currently uses historical experience; and claims can be messy processes of subjective reminiscences. Autonomous cars, on the other hand, will be able to record, monitor and save all the details of any given journey.
This twin shift of data and liability will require insurers to reconsider their product offering. The concept of driver responsibility will be eroded; and product liability will be enhanced. Using driver data, insurers could split their product offering between product and driver liability, depending on who is driving and for how long. This split could be covered under a single product with separate pricing structures; or could be divided, for example the car manufacturer assuming product liability, and the individual driver the manual driving portion. There could even be some form of pay-as-you-go insurance, based on distance driven.
There are many potential permutations. What is clear is that as driving becomes more data reliant, insurers will have to become more flexible. Successful insurers will be those engaging with this data rich environment so that insurance products mirror the coming changes in our mobility.
The eventual coming of fully autonomous vehicles offers new challenges for insurers. Liability will fall entirely on the product, which some believe will be so safe as to barely warrant any coverage. Already there have been articles suggesting 'the end of car insurance.'
We remain someway from that point. However, insurers can begin to build potential scenarios, such as that below:
MTPL – Motor Third Party Liability
MOD – Motor Own Damage
GTPL – General Third Party Liability
Technology is currently at the front of the development of the autonomous car. It is, however, far from the only factor that will determine how quickly we will move along the road to the robot car. Many societal factors will play a role in the adoption of new technology and how quickly insurers will have to adapt their product range:
- Regulatory authorities: Some authorities, such as Singapore, Gothenburg and California, have been keen to encourage the development of autonomous technology on their roads. These authorities could see their states and cities become early adopters of the autonomous car.
- Pressures of urbanisation: Autonomous car technology could be seen as particularly attractive to those cities suffering chronic congestion. Those authorities in need of radical mobility solutions may watch the development of autonomous cars with particular interest.
- Consumer preference: A number of consumer groups will benefit from greater vehicular autonomy. For the elderly (a growing constituency) and the disabled, there is the prospect of greater mobility. Time-poor workers may use autonomous cars as ersatz offices. Those with long commutes may simply alleviate the boredom with in-car entertainment.
- Generation shifts: There is some debate as to whether younger generations strive to own their own car, in the way their parents did. A number of countries already have car sharing schemes, which may appeal more to younger drivers, particularly when education and housing costs are rising faster than potential incomes.
- Cost: Moore's Law is at work across digital technologies. The sensor and radar equipment making current hybrid cars more expensive than entirely manual models will fall dramatically in price, so that they will become part of the standard product offer very quickly.
- Regulatory authorities: Some regulatory authorities will make life harder for autonomous cars. Those with already good public transport may see less urgency in making adjustments that facilitate the development of driverless cars.
- Legal systems: As noted in an article in this publication, legal systems can adapt to changing realities; but that change can be slow. Particular legal issues for autonomous cars are likely to be criminal liability and data protection.
- Consumer preference: There will be consumers who do not like autonomous cars. Many drivers enjoy manually driving. Autonomous cars are likely to have a more conservative bias, which will annoy risk takers and faster drivers. Other consumers may simply not trust autonomous technology.
- Cost: Even if technology costs fall, cars without autonomous functions will be cheaper to produce for some years yet. Some manufacturers may even offer basic models with no autonomous features. The longer manual cars are still on the road, the longer it will take to realise the full potential safety gains of autonomous vehicles.
- Political resistance: There are many scenarios in which politicians and the public may turn against autonomous cars. One bad accident with an autonomous car, say crashing into a school bus queue, could set the technology back years. There are vested interests in driving professions, such as taxi drivers, haulage contractors etc., who may not welcome fully autonomous cars.
Three broad conclusions can be drawn from this edition of the Risk Dialogue Magazine:
1. Autonomous technology is already here and will be further developed. All major car producers plus prominent outsiders are investing millions in car computerisation. Insurers should already be positioning themselves for incremental increases in autonomous technology, which first and foremost will be interfacing with new data sources.
2. The speed of future development of autonomous technology and the fully autonomous car will depend on a number of non-technical factors. Insurers will also need to pay attention to the courts, consumer preference and a number of other factors.
3. The further we look into the future, the hazier the vision of the autonomous car becomes. We may inhabit a world of robot-driven taxis, with drivers no longer required. It is not inevitable. There have been previous promised revolutions in mobility that did not occur. However, while the potential is there, it is something insurers should plan into longer term scenarios.
Download full article (PDF 320 KB)