Selim Cavanagh, Vice President, LexisNexis Risk Solutions
Autonomous vehicles (AVs) and other new types of mobility will have a material impact on society. It is becoming increasingly clear how telematics, and associated connected car technology, are laying the foundations for the driverless car ecosystem, so doing nothing is no longer an option for insurers.
Here are some of the important forecasts. Tesla© has so far completed close to 250 million miles of motorway driving using its Autopilot but it says it needs “billions” of miles before full driverless technology (Level 5 on the SAE scale) can go to market. Google’s Waymo also wants to accelerate its coverage of road miles: it has approximately 2 million miles under test so far, including 700,000 miles in 2016.
In the automotive world, everyone wants to move as fast as possible to a Level 5 autonomous vehicle and Tesla recently moved forward its target launch for this vehicle type to 2018.
The expectation is that safety – and frequency of accidents and claims – will improve with autonomous features, demanding a response from the insurance industry in terms of risk scoring and data validation.
This will happen around the world, even if for now regulators will not allow driving with “no driver present”, which is the Level 6 vehicle type.
Uber began to test its AV technology (with driver and engineer) in Ford cars in September 2016 in Pittsburgh. It expects to have a fully driverless fleet by 2030. Also last year General Motors and ride hailing company Lyft™ announced a deal to develop an AV fleet. Lyft expects the release of full AVs to be in 2020 and it estimates that some 80% of its fleet will be autonomous by 2022.
Google can reach one billion miles travelled by 2025‐2026 if it doubles its mileage each year. Data‐gathering from the behaviour of hybrid (less than fully autonomous) vehicles will likely be needed to augment the test mileage, to understand the full impact of ADAS (Advanced Driver Assistance Systems) and measure how driver behaviour – and pedestrian behaviour – will change in the future.
Forecasts from Ptolemus predict that by 2025, 100 million vehicles will be on the road with some form of advanced cruise control in the US and Europe. By 2030, there will be more cars on the road globally with ADAS than without. Specifically, by then 370 million vehicles will have some automated features and 13 million will be highly autonomous.
The research firm has defined vehicle autonomy as potentially a “perfect storm” for insurance, ready to wipe out risk. Most automotive OEMs will not launch a Level 3 vehicle (the type capable of autonomously steering, braking and changing lanes), preferring to jump straight to Level 4 (full autonomy but with a driver) to avoid re-engagement and liability issues.
Mobility taking over from vehicle ownership?
Driverless (shared) cars are likely to arrive on the market before Level 4 (owned) cars because they will be introduced in restricted areas, away from public roads, first.
As we explored in a previous blog, depending on how this aspect of the market plays out, the OEMs will shift to become mobility service providers, engaging with car clubs and ride hailing firms, as a means to retain the customer relationship, and offset the risks of decreased car shipments.
Up to one in ten cars sold in 2030 could potentially be a shared vehicle, with implications for insurance products and the whole value chain.
In general the third party forecasts from the likes of KPMG, Deloitte and McKinsey & Company are predicting availability of autonomous vehicles five to ten years later than the vehicle manufacturers. This is based on the high cost of the technology, as AV capability will add nearly £10,000 to the cost of a vehicle, and their view of public perceptions and the regulatory timescale.
But what do we make of questions around risk and liability, considering that with autopilot in the aviation world, recent precedents point to an element of both human error and manufacturer liability when considering recent airplane accidents?
Solving questions around liability, based on accepted proof points of data, will become very important for insurers and the vehicle OEMs, to be able to establish public confidence and adoption by consumers.
So how do UK motor insurers view the prospect of a driverless world, or what really amounts to a hybrid driven/driverless world for the next 15 years or so?
In a recent survey* we asked motor insurers about attitudes to a range of digital offerings in insurance, including 12V plug-in telematics devices, data prefill of applications, peer-to-peer insurance, social broking and autonomous vehicles.
Motor insurers expect that 12V telematics devices and autonomous vehicles will have the greatest impact on the industry. But only a small segment expressed this with strong certainty. Insurers were more mixed on whether they felt peer-to-peer insurance and social brokers will reach the mainstream.
Here are some of the important forecasts. Tesla© has so far completed close to 250 million miles of motorway driving using its Autopilot but it says it needs “billions” of miles before full driverless technology (Level 5 on the SAE scale) can go to market. Google’s Waymo also wants to accelerate its coverage of road miles: it has approximately 2 million miles under test so far, including 700,000 miles in 2016.
In the automotive world, everyone wants to move as fast as possible to a Level 5 autonomous vehicle and Tesla recently moved forward its target launch for this vehicle type to 2018.
The expectation is that safety – and frequency of accidents and claims – will improve with autonomous features, demanding a response from the insurance industry in terms of risk scoring and data validation.
This will happen around the world, even if for now regulators will not allow driving with “no driver present”, which is the Level 6 vehicle type.
Uber began to test its AV technology (with driver and engineer) in Ford cars in September 2016 in Pittsburgh. It expects to have a fully driverless fleet by 2030. Also last year General Motors and ride hailing company Lyft™ announced a deal to develop an AV fleet. Lyft expects the release of full AVs to be in 2020 and it estimates that some 80% of its fleet will be autonomous by 2022.
Google can reach one billion miles travelled by 2025‐2026 if it doubles its mileage each year. Data‐gathering from the behaviour of hybrid (less than fully autonomous) vehicles will likely be needed to augment the test mileage, to understand the full impact of ADAS (Advanced Driver Assistance Systems) and measure how driver behaviour – and pedestrian behaviour – will change in the future.
Forecasts from Ptolemus predict that by 2025, 100 million vehicles will be on the road with some form of advanced cruise control in the US and Europe. By 2030, there will be more cars on the road globally with ADAS than without. Specifically, by then 370 million vehicles will have some automated features and 13 million will be highly autonomous.
The research firm has defined vehicle autonomy as potentially a “perfect storm” for insurance, ready to wipe out risk. Most automotive OEMs will not launch a Level 3 vehicle (the type capable of autonomously steering, braking and changing lanes), preferring to jump straight to Level 4 (full autonomy but with a driver) to avoid re-engagement and liability issues.
Mobility taking over from vehicle ownership?
Driverless (shared) cars are likely to arrive on the market before Level 4 (owned) cars because they will be introduced in restricted areas, away from public roads, first.
As we explored in a previous blog, depending on how this aspect of the market plays out, the OEMs will shift to become mobility service providers, engaging with car clubs and ride hailing firms, as a means to retain the customer relationship, and offset the risks of decreased car shipments.
Up to one in ten cars sold in 2030 could potentially be a shared vehicle, with implications for insurance products and the whole value chain.
In general the third party forecasts from the likes of KPMG, Deloitte and McKinsey & Company are predicting availability of autonomous vehicles five to ten years later than the vehicle manufacturers. This is based on the high cost of the technology, as AV capability will add nearly £10,000 to the cost of a vehicle, and their view of public perceptions and the regulatory timescale.
But what do we make of questions around risk and liability, considering that with autopilot in the aviation world, recent precedents point to an element of both human error and manufacturer liability when considering recent airplane accidents?
Solving questions around liability, based on accepted proof points of data, will become very important for insurers and the vehicle OEMs, to be able to establish public confidence and adoption by consumers.
So how do UK motor insurers view the prospect of a driverless world, or what really amounts to a hybrid driven/driverless world for the next 15 years or so?
In a recent survey* we asked motor insurers about attitudes to a range of digital offerings in insurance, including 12V plug-in telematics devices, data prefill of applications, peer-to-peer insurance, social broking and autonomous vehicles.
Motor insurers expect that 12V telematics devices and autonomous vehicles will have the greatest impact on the industry. But only a small segment expressed this with strong certainty. Insurers were more mixed on whether they felt peer-to-peer insurance and social brokers will reach the mainstream.
- 66% of motor insurers believe 12V telematics devices will definitely/probably have a positive impact on insurance
- 64% of motor insurers believe autonomous vehicles will definitely/probably have a positive impact on insurance
- 54% of motor insurers believe data prefill will definitely/probably have a positive impact on insurance
- 49% of motor insurers believe peer-to-peer insurance will definitely/probably have a positive impact on insurance
- 44% of motor insurers believe social broking will definitely/probably have a positive impact on insurance.
In terms of the digital challenges generally, 45% of insurers across motor, home and commercial said data protection and security are a top concern, followed by easier switching by consumers (44%), social media complaints (44%), cyber risk (39%) and payment fraud (39%).
We believe that AVs and digital automotive technology will tend to drive a wedge between two groups of insurers: those that are able to attract and price good risk – the self-selected drivers who by nature are safer and willing to be monitored – and those insurers that are left with an increasingly long tail of poor risk.
While projections vary, it seems clear that the next three to five years will witness:
- Increasing consumer awareness and demand for telematics and connected car services amongst broader driver types
- Decreasing telematics hardware cost, corresponding to much larger data samples, lower-cost of data and increasing granularity of data for further improvements in risk scoring
- Increasing clarity about the parameters for new services such as driver alerts, first notice of loss (FNOL) standards and telematics data in claims handling
- The introduction of first-generation telematics products by many more insurers (and second and later-generation products by others)
- Competitive gains for insurers with leading telematics products at the expense of rivals.
The pace of change is accelerating for insurance telematics, and the insurers that are acting now will be best positioned to attract desirable policyholders for years to come.
*LexisNexis Risk Solutions carried out an anonymous survey, the UK Insurance Underwriting Digitization Study, 8 December 2016–9 January 2017. Mixed mode of data collection: online panel and telephone interviewing. The sample was 170 insurance professionals, 55 personal motor, 52 personal home and 63 commercial property. The respondent must spend 30% of their time in underwriting-related activities for a given line to be assigned to answer questions specific to that insurance line.
This article is care Lexisnexis and the original article can be found here
