10.4 - The Autonomous Vehicle Flashcards
Describe the autonomous vehicle and the impact it will have on the insurance industry
Vehicle Automation
-vehicle automation can produce a significant reduction in traffic fatalities and serious injuries
-the goal is to reduce the number of collisions as more vehicles include new safety technology and systems, such as forward-collision avoidance
-as the capability of technical equipment improves, the societal benefits are likely to increase as well
Artificial Intelligence and Vehicle Automation
Artificial Intelligence (AI) is a property of machines, in contrast with the natural intelligence of humans and other animals
-the term ‘artificial intelligence’ was coined by the American computer scientist John McCarthy in 1955, and the academic discipline dates from the Dartmouth Workshop that McCarthy proposed and helped organize the following year at Dartmouth College in New Hampshire
-AI combines computer science and robust datasets to enable problem solving
-early research led to expert systems that tried to imitate human decision making
-today the focuses of research are machine learning and deep learning, which involve algorithms to predict or classify based on input data
-machine learning has improved some expert systems and made them easier to create
-AI supports many features of life in the modern world, including Internet search engines, product recommendations, and speech recognition systems
-AI will be the technology that allows vehicles to eventually become fully autonomous
-furthermore, AI will have many other applications for insurer and companies, such as strategic decision making, risk analysis (underwriting), predicting claims frequency and severity, and assisting with fraud detection and prevention
Semi-Autonomous or Self-Driving Vehicles
-horse and buggy first “self-driving” vehicle in Canada 150 yrs ago
-efforts to create self-driving cars typically failed due to limitations in technology available at the time and the cost of implementation
-self-driving vehicles are now being tested on public roads
-the first fully automated, self-driving cars, supported by sensor technology and on-board computing, could be available to consumers within the next 5-10 yrs
-AI will drive this new technological innovation
-as AI improves over time, it will be able to simulate human decision making and replace the human driver
-over the long term, self-driving vehicles hold great promise to transform society, including dramatically reducing traffic fatalities and serious injuries
-Automakers have promised driving-assistance systems over the next 5-10 yrs that will automate some aspects of driving semi-autonomous vehicles
-for example, cars with intelligent cruise control and lane-keeping mechanisms would allow drivers to travel on a highway without putting their hands on the steering wheel or using the brakes and throttle
-cars with a forward collision avoidance system could prevent collisions with other vehicles or objects that are ahead
-vehicles with auto-park would park themselves even if the driver gets out to watch
-semi-autonomous vehicles with some of these features are available today and hold great promise to reduce traffic collisions and injuries
Responsibility for Collisions
-responsibility for collisions will begin to shift from drivers to vehicles, parts manufacturers, and software and sensor providers
-in the near future, roads will be shared by fully-human driven vehicles, semi-autonomous vehicles with driving assistance systems, and the first self-driving vehicles
-as on-board computers and AI begin to make driving decisions, responsibility for collisions will move beyond human drivers to include automakers, software developers, and maintenance professionals
-there may be a shift from personal liability for collisions involving conventional vehicles to shared liability for semi-autonomous vehicles, and predominantly product liability for self-driving vehicles
Levels of Automation
-in 2021, the Society of Automotive Engineers (SAE) International issued its Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles (commonly referred to as SAE Levels of Driving Automation)
-this document provides a consistent way to describe the range of automation features that may be included on motor vehicles
6 Levels of Automation:
> Level 0 - No driving automation: the vehicle has no driving automation technology. Full time performance by the human driver of all aspects of the dynamic driving task is required, even when enhanced by warning or intervention systems, such as forward-collision warning, automated emergency breaking, blind-spot warning, and lane0-keeping assistance
> Level 1 - Driver Assistance: The vehicle has at least 1 driver-support system to provide steering assistance or acceleration/deceleration assistance, using information about the driving environment and with the expectation that the human driver performs all remaining aspects of the dynamic driving task, such as adaptive cruise control
> Level 2 - Partial driving automation: The vehicle has advances driving-assistance systems that control both steering and acceleration/deceleration, using information about the driving environment and with the expectation that the human driver performs all remaining aspects of the dynamic driving task
> Level 3 - Conditional driving automation: The vehicle has advanced driving-assistance systems and AI that make decisions based on changing driving situations surrounding the vehicle and all aspects of the dynamic driving task, with the expectation that the human driver will respond appropriately to a request to intervene (physically or remotely), such as in an emergency due to a system failure
> Level 4 - High driving automation: The vehicle has an automated driving system for all aspects of the dynamic driving task. It is programmed to stop itself in the event of a system failure with no expectation for a human driver to intervene (for example, driverless taxis)
> Level 5 - Full Driving automation: The vehicle has an automated driving system for full-time performance of all aspects of the dynamic driving task under all roadway and environmental conditions that can be managed by a human driver
-Level 0 vehicles are described as conventional vehicles
-Level 1 and 2 vehicles are described as semi-autonomous
-the driver is responsible for the safe operation of the vehicle at all times but has access to support in some situations
-with supervision by the driver, these vehicles may be capable of driving themselves in some circumstances, like following a car in stop-and-go traffic or parking
-Level 3, 4 and 5 vehicles are described as self-driving cars and trucks
-these vehicles are capable of fully driving themselves in most situations, but drivers may need to assume control in some circumstances
Implications of Autonomous Vehicles for the Insurance Industry
-important changes for the insurance industry will be most evident when self-driving cars become more commonplace
-current insurance coverages and practices are not designed for a world where vehicles that drive themselves replace human drivers
-SAE Level 0 to 2 vehicles are expected to account for most of the vehicles on Canadian roads over the next 10 years
-it is unclear how the introduction of SAE vehicle Levels 3 to 5 (including fully self-driving vehicles) will change the collision experience of conventional/partially automated vehicles (those vehicles where the human driver is still tasked with monitoring the environment)
-it is also unclear how responsibility will be determined when conventional vehicles collide with partially automated and fully automated vehicles
-IBC releases a proposal for Canadian automobile insurers in 2018 to help guide them in determining responsibility in these cases
-IBC proposed “a single insurance policy that covers both driver negligence and the automated technology”, permitting the auto insurer to indemnify an innocent TP regardless of the cause of loss and to recuperate its losses by subrogating against the vehicle manufacturer or technology provider if warranted
-the insurance industry will face several challenges as it provides coverage for new vehicles with driving-assistance technology
-automakers are promising a marked reduction in the risk of collisions in semi-autonomous vehicles, but insurers will have little initial evidence about the frequency and severity of claims
Questions about semi-autonomous vehicles emerging for the industry
-how can the insurance industry secure timely information about the collision experience and repair costs for semi-autonomous vehicles?
-will automakers install a “black box” to record when driving-assistance features are engaged?
-will insurance companies be allowed to access this data?
-what safety technology will be required by regulation in new vehicles?
-will the coverage offered need to be redesigned?
Questions about self-driving vehicles emerging for the industry
-will the insurance coverage for the first self-driving vehicles be modelled on the product liability coverage currently in place for aircrafts, ships, and trains with substantial automation?
-what incremental coverage will be offered to drivers who will have the option to take control of self-driving vehicles?
-will auto rate regulation apply to product liability coverage for the first self-driving vehicles?
-will insurance cover a variety of vehicle ownership alternatives like personal ownership, carsharing, ridesharing, ride hailing, and pooled ride hailing?
-is the current construction of insurance coverage appropriate for vehicles with automation?
-what decisions will manufacturers and regulators make over the next decade that could determine the nature of connected vehicles over the long term?
-will the definition of driver be changes to include both physical and remote operators? How will this impact acceptability and underwriting rules?
Implications of Artificial Intelligence for the Insurance Industry
-online chat feature or quoting an insurance policy online are basic examples of AI at work
-insurers and companies are only just starting to use AI in their operations, and companies have only begun to explore and develop the applications and use of AI
-some expanded uses of AI that the property and casualty industry will likely implement in the near future include the following:
>risk analysis and pricing in the underwriting process
>predicting the frequency, severity, or development of claims
>helping identify fraud and loss preventions
