Autonomous Vehicles implications for insurance Scott Reeves, Munich Re Image from Google images
Agenda 1. What we know so far 2. What s over the horizon? 3. Near-term implications for the insurance industry? 4. How will the insurance industry respond? 2
I have already been in a vehicle where I had almost no control 3
What we know so far
Automation is not just one thing NHTSA levels of automation Full Automation Level 4 Testing Limited Automation Level 3 Testing Combined Function Level 2 In production Function-specific Level 1 In production No-Automation Level 0 No longer in production The vehicle is self-driving for entire trip; no human control required Driving functions are sufficiently automated under certain traffic or environmental conditions; the driver can safely engage other activities but expected to be available for occasional control. (Google car) More than one function is automated at the time, but the driver remains attentive. Example: adaptive cruise control in combination with lane centering One or more functions is automated, but under human control. Examples: electronic stability control or pre-charged brakes The human driver is in complete control of all functions of the car Source: National Highway Traffic Safety Administration, Preliminary Statement of Policy Concerning Automated Vehicles, 5/2013 5
Understanding autonomous vehicles Sensors, cameras, radar and Light Detection and Ranging (LIDAR) provide data about vehicle s position and surroundings to a computer that controls the car. LIDAR, like radar, fires bursts of energy at a target and measures the return time to calculate the distance. LIDAR generates high-resolution, threedimensional data. Image: Google, Inc. Google and the Google logo are registered trademarks of Google Inc., used with permission. 6
Human vision vs. machine vision Field of view: 100 degrees Distance: 30 meters ahead Field of view: 360 degrees Distance: 300 meters ahead 7
Status of AV technology Number of unique inventions v Google ranked 19 th for AV patents Source: Thomson Reuters, The 2016 State of Self-Driving Automotive Innovation, 1/7/2016 8
Status of AV technology car manufacturers Manufacturer Target Model/Focus Audi 2017 Audi to pilot driving A8 in 2017 BMW 2015 Prototype partnership with Chinese search engine company Baidu Ford 2020 Smart Mobility plan may produce car within five years GM 2020-2030 Semi-autonomous by 2020; fully autonomous by 2030 Mercedes 2020 2015 Fully autonomous E, S Class models imminent Daimler Freightliner Inspiration truck licensed in Nevada Nissan 2020 Partnership with NASA to progressively introduce AV technologies to consumers beginning in 2016 up to 2020 Tesla 2015 Autopilot Model S sedan introduced in 2015 Toyota/Lexus Unspecified Partnership with Stanford University and Massachusetts Institute of Technology to develop "human-centric systems Volvo 2017 100 fully autonomous on road in Sweden 2017 Continental component / retrofit 2016 2025 Low-speed in traffic Fully-autonomous Source: Manufacturer s websites, 2016 9
Current success stories Delphi: coast to coast (5,500km) 99% AV Google: over 1m miles driven* Source: CNNMoney, Driverless Car Finishes 3,400 Mile Cross-Country Trip, 4/3/2015, Google, Self-Driving Car Project, Accessed 3/7/2016 Image source: Delphi, Google 10
The AV revolution has not really begun yet Source: McKinsey & Company, Ten Ways Autonomous Driving Could Redefine the Automotive World, 6/2015. 11
What s over the horizon?
Ready to take you where you (might) want to go Based on your web browsing before you walk out the front door, the car will assess the most likely place you want to go first Source: Google images 13
Protection against natural hazards Avoid hail by advance-alert Enact disaster response plan Go to pre-identified undercover site (but how far away now?) Do you have control? Source: Google images 14
No need for large urban parking stations Parking stations will not be required your AV will drop you at the door, and then Drive home, or Drive to a designated parking lot many km away (in wasteland), or Convert to a Uber taxi Source: Google images 15
Car ownership may fall Today your car is an unused asset 95% of the day. Will we own fewer cars if they are available ondemand? Will you hire-out your car? Or send it on errands when it s not in use? Source: Forbes, Self-Driving Cars Are Coming, 10/13/2014 16
Livery options Google and Uber compete to develop driverless taxis. Eliminating the driver significantly reduces overhead costs. Consumers can expect a driverless Uber fleet by 2030. Source: Mobility Lab, Uber s Plan for Self-Driving Cars Bigger Than It s Taxi Disruption, 8/18/2015 17
Near-term implications for insurance
There is enormous scope for improvement Driver error is the cause of 94% of accidents In 2015 in the US* $412** billion accidents 4.4 million injuries 38,300 fatalities *provisional estimate increase of 8% year over year, largest increase in 50 years **cost of motor-vehicle deaths, injuries and property damage Source: National Safety Council Vehicle Fatality Estimates, 2/17/2016 19
We are already seeing significant safety improvements* KEY: LDW - Lane departure warning ACC - Adaptive cruise control Source: Insurance Institute for Highway Safety (IIHS), 2015 20
Insurance industry will lose business As of 2015, approximately 38% of general insurance industry premium is derived from motor vehicle insurance According to Celent, US auto insurance premiums could drop 60% starting in 2020 s IHS Automotive predicts worldwide self driving vehicles sales of 230,000 by 2025; 11.8m by 2035 Source: APRA Net Earned Premium (Domestic MV, CTP) 31/12/2015,, SNL, IHS Automotive, Emerging Technologies: Autonomous Cars Not If, But When, 1/2/2014, Celent, End of Auto Insurance, 5/2012 21
But zero crashes are not guaranteed In the real world there are always unknown moving obstacles... it s always possible to find situations where a collision will happen. Thierry Fraichard Project-Team Perception, recognition and integration for modelling of activity (PRIMA) 2014 As the Google AV was reentering the center of the lane it made contact with the side of the bus. The Google AV was operating in autonomous mode and traveling at less than 2 mph, and the bus was traveling at about 15 mph at the time of contact. California Department of Motor Vehicles, Google Accident Report 2/14/2016 22
What is the cost of repair? AV technology is not just one system Source: BCG, Revolution in the Driver s Seat: The Road to Autonomous Vehicles, 4/21/2015 23
Is the insurance industry doomed as we know it? Self-driving cars and ride-sharing programs will completely disrupt the carinsurance industry. "Accident frequency will decline to where the difference among driving behaviors becomes negligible and it is difficult to charge a meaningful premium for insurance." "Insurance will take the form of commercial product liability instead of personal driver liability as we let the robots do the driving." "Vehicle utilization will rise and cars on the road will decline as one car can serve the driving needs of multiple travelers per day, which, in-turn, means fewer cars." Business Insider This could be the beginning of the end for the car-insurance business July 21, 2015 24
Likely implications for insurance premiums CTP Cyber risk tech E&O/IoT Likely to shrink Likely to increase Products liability Transition to full vehicle autonomy ------ Varying degrees of impact over time Comprehensive Motor Likely no material change Equipment breakdown/warranty Likely to increase Product recall Likely to increase Likely to increase 25
Products potential coverage implications Expertise - Potential difference between the AV company and existing auto parts/technology is the amount of knowledge that the auto manufacturer has about the product or technology used. Causation What caused the loss/accident? Will parts manufacturers exposure be evident in all cases? Allocation/Accountability Who is responsible? Given that the AV company is providing specs to the OEM (and not vice versa), is the AV company ultimately liable if the auto is built to spec, even if the part fails? Reputational risk - Given the amount of media attention surrounding this technology, if there are issues, does this technology present a reputational risk to the AV company s brand and other businesses? AV manufacturer may pick up liability - e.g. Volvo, Mercedes, Google.Who next? 26
Liability issues AV cyber vulnerability Vulnerable technology and software Product liability technology E&O Traditional Bodily injury / Property damage Privacy Equipment malfunction Hacking / Malfeasance Example: Jeep 2015 hackers (researchers) accessed a Jeep while it was driving Took control of the radio, wiper blades and brakes Used a laptop and Smartphone to hack the Jeep from 10 miles away Image - https://commons.wikimedia.org/wiki/file:jeep_grand_cherokee_3.0_crd_overland_ (WK)_%E2%80%93_Frontansicht_(1),_10._September_2011,_D%C3%BCsseldorf.jpg 27
How will the insurance industry respond?
Looking ahead Inevitable Complex Evolving Challenging Insurance of motor vehicle exposure will change, possibly dramatically over time Liability shifts from driver to manufacturers and technology companies Comprehensive motor, cyber, products warranty may grow Exposures will be more complex Cyber/software, car manufacturer, driver Coverage issues will emerge and take time to evolve and stabilize Telematics use will grow continue the progress towards individual ratemaking Insurance industry should understand the issues; be prepared to adjust and innovate (particularly those with material personal or commercial portfolios) 29
The matter of liability will evolve Strict and fault-based liability will continue to apply: victims will need to be fully compensated, regardless of cause! Continued coverage through policies, with strict liability also applying for driverless robotic vehicles. Possible consequences for product liability if motor insurers seek recourse from manufacturers. For consideration: Was the accident caused by failure of a technical unit or a software error, or was it caused by a driver error? Altered liability; the autonomous functionality will lead to an increase in the manufacturers liability. In Germany, vehicle owners will continue to be liable (including for service and software updates), but they must not be considered test drivers of immature systems or have to prove product failure. Insurers must retain the right to seek recourse from manufacturers. In the US, where coverage levels are often lower, the focus on product liability is likely to be greater in future. New responsibilities will arise for manufacturers, suppliers and third parties under product liability laws; users will still have to intervene for a relatively long time and thus continue to bear responsibility. Besides: Human or machine who will ultimately have responsibility? 30
Altered risks for the insurance industry Rapid but still unclear development of technology and local law / court ruling, different national regulatory regimes Complex technology Complex liability issues Increased claims expenditure incurred by motor vehicle insurers owing to complex liability issues; specialists with high technical expertise needed Insurers need to understand behaviour of the new systems in order to assess the risks, i.e. deep understanding of risks technical know-how Continuous need for new risk assessments by insurers and support from reinsurers German multi-line research project acatech: Need for cross-border cooperation between various actors involved in automated driving: R&D, supply side and users. Science and industry should develop uniform principles of human-machine interaction. Partnering with customers, manufacturers and technology developers (telematics, telecommunication) increasingly important. Munich Re partnerships, e.g. with Comet (feasibility study) and with scientific institutions such as Mcity (University of Michigan), initiatives such as Mobility Domain 31
Stakeholders in the outcome Often mentioned Drivers AV Manufacturers Software providers Insurers Panel repairers But also Injured people and families Governments Health care sectors 32
Product insurance is not necessarily the end state dependence on AV functionality less fault from driver decision being taken from driver? Focus on motor insurance Focus on product insurance Complexity of establishing fault Government budget pressure Care / indemnity balance 33
Specific issues for people injured in accidents Will CTP as a product continue? A Products Liability regime may see: Delays to settlement, no interim payments Uncertainty of available limit Claimants to follow a class action? CTP provides funding, taking pressure from the injured person, and government budgets. Likely the NIIS provides care for injured people, funded from the states Might state governments co-opt existing CTP arrangements to apply to your responsibility for consequences of the vehicle. 34
Munich Re in partnerships Munich Re is a supporting partner of the Australian Driverless Vehicle Initiative We are a member of the Policy and Risk Group, which has responsibility for planning, development and coordination of all position papers, policy and risk forums Other insurance partners are IAG, Suncorp, and Zurich 35
Munich Re - a thought leader in the AV space Identifying promising tech start-ups in Silicon Valley Partnering with AV start-ups to stimulate progress Researching AV emerging risks Collecting and analyzing data Developing innovative products and solutions Educating and empowering stakeholders 36
Disclaimer This presentation contains forward-looking statements that are based on current assumptions and forecasts of the management of Munich Re. Known and unknown risks, uncertainties and other factors could lead to material differences between the forward-looking statements given here and the actual development, in particular the results, financial situation and performance of our Company. The Company assumes no liability to update these forward-looking statements or to make them conform to future events or developments. 37
Images: depositfotos.com Thank you very much for your attention. Scott Reeves 2016 Münchener Rückversicherungs-Gesellschaft 2016 Munich Reinsurance Company