ADAS(高度ドライバーアシスタントシステム）への道: 自動車保険保険料設定におけるADASデータの利用

Ok, full Level 5 autonomy is receding in the horizon, no matter how hard we press on the automotive technology accelerator.

What we do have today, and will increasingly have in the next few years, is ADAS. ADAS is the unlovely, but widely used, acronym for Advanced Driver Assistance Systems. I personally would have preferred something like “Onboard Collision Avoidance Technology,” but alas “ADAS” is in common usage.

ADAS is the umbrella term for a broad set of technologies, built into a growing proportion of new vehicles, that make collisions less likely, and possibly less severe. Some ADAS features are passive, in that they only warn a driver of an impending crash (or inherently risky driving behavior). Other ADAS features are active, meaning that they will take control of some aspect of the vehicle’s operation (typically braking or steering). Examples include:

• Passive: Lane Departure Warning, Forward Collision Warning, Rearview Cameras
• Active: Automatic Emergency Braking, Adaptive Cruise Control (the vehicle maintains a given distance a vehicle traveling ahead)

The intent of ADAS technologies, from an insurance perspective, is laudable. But what really matters is loss data: frequency and severity by coverage, by make and model, and by type(s) of ADAS technologies. The Insurance Institute for Highway Safety (IIHS) and its sister organization, the Highway Loss Data Institute (HLDI), have published important studies analyzing loss data for specific makes, models, and ADAS equipment over a period of years – here, here, and here. In general these studies have provided encouraging evidence regarding ADAS effectiveness.

However, what has been missing is a way for insurers to correlate loss experience for all their insured vehicles to the presence or absence or ADAS technologies in those vehicles. VIN numbers do not indicate ADAS equipment on a specific vehicle. And policyholders’ self-reports have certain reliability issues.

This situation is now changing.

• In March 2019, ISO introduced a new manual rule and related rating factors for personal auto. It also introduced a data set, VINMASTER® –VINterpret Safety, which indicates whether a vehicle has certain ADAS features as a standard or optional feature.
• In June this year, LexisNexis launched its Vehicle Build program. Vehicle Build provides ADAS status information specific to individual vehicles, and groups that information into a proprietary LexisNexis ADAS Classification System

The proportion of ADAS equipped vehicles in the fleet on the road is growing steadily. A LexisNexis study found 76% of 2019 models with one or more ADAS feature – a substantial increase from 18% in 2014 models.

Data sets that provide granular information about ADAS equipment will accelerative insurers’ journey on the road to ADAS use in rating algorithms.