Dec 18 2024
The auto insurance industry, after several years of experiencing unprecedented increases in loss ratio, managed to stem the tide in 2023 by posting a 104.9% mark, a 7% improvement from the prior year that still remains as a net loss. As these organizations continue to see business in the red, owner behavior is shifting in a manner that could easily see a return to loss increases. According to LexisNexis, high claim severity persists, and increased EV sales (up 54% year-over-year) could also inflate loss due to their increased susceptibility to total loss.
Vehicle technology is increasing in complexity besides EVs, and that complexity is a prime driver in added repair and claim costs. A recent AAA report that repairs focused on Advanced Driver Assistance Systems (ADAS) make up over 37.6% of repair costs after a crash. The industry-wide trend toward full vehicle automation is also fundamentally shifting risk from the driver to the car, and insurance risk modeling has not yet caught up to satisfactorily factor this in. McKinsey writes that this “connected revolution” in the automotive industry “means insurance coverage will likely need to shift from drivers to the automakers and software companies responsible for the development and maintenance of various autonomous-driving technologies.”
A fundamental problem with this vehicle-centric approach to auto insurance risk analysis is that OEMs do not describe the same ADAS feature sets in the same manner. Capabilities for the same branded ADAS features can even change between the same vehicle’s model years, as shown with Acura’s updates to its AcuraWatch system. Understanding which ADAS systems are actually installed on a vehicle is a prerequisite to modeling risk, but a lack of standardized terminology used in a consistent manner across OEMs makes this very difficult.
Additionally, since different types of sensors and assistance packages vary dramatically in capability, this can create a significant degree of difficulty in weighing vehicle risk. Characteristics such as the minimum and maximum speeds and distances systems operate greatly impact their ability to mitigate collisions. And, factors like the location and quantity of sensors also create large variances in estimated repair costs – for example, are the ADAS sensors located in the bumper, prone to damage in even low-speed collisions, or behind the windshield? How many ultrasonic sensors are installed in the vehicle? Some systems use only two while others may use six or more. All of these variables mean that a surface-level ADAS feature description is not sufficient to accurately assess risk in policy generation and later claim judgment.
DataOne Software’s vehicle datasets have been consistently informed by the automotive industry’s evolving technology challenges. Our latest vehicle data solution streamlines identifying the granular components of ADAS packages to power the next generation of insurance risk rating models...
DataOne’s ADAS dataset includes standardized definitions for the components of automotive driver assistance system packages across all brands and model years, beginning with the advent of modern ADAS technology in model year 2015. Users of this data also benefit from knowing the location and placement of the sensors and assistive equipment on vehicles, to better estimate the likelihood of damage in a collision and repair costs.
Beyond supplementing the marketing terminology attached to these feature sets with standardized terminology, DataOne’s ADAS data solution also accounts for the standard operating parameters for these systems (such as minimum and maximum speeds and distances, and the driver’s ability to manually configure each feature). This data gives insurance organizations straightforward, highly detailed information to power risk modeling and claims processing. DataOne’s latest ADAS data release can also be tied into OEM build data and verified records, to provide insight into the optional ADAS features that consumers can select, ensuring a clear picture of the vehicle’s assistive capabilities.
If you are interested in learning more about DataOne’s ADAS data product, you can contact a data strategist at the button below.
