Objective:​ Analysis examined how individual collision avoidance features affected losses under a variety of insurance coverages for vehicle damage and injuries.

Methods:​ Five automakers supplied identification numbers of vehicles that had each feature, allowing a comparison of the insurance records for those vehicles that included the optional feature with the same models without the feature. Coverage and loss data were supplied by insurers representing over 80 percent of the U.S. private passenger vehicle insurance market. Regression analysis was used to quantify the effect of each vehicle feature while controlling for the other features and covariates, including driver age and gender, garaging state, and collision deductible. Claim frequency was modeled using a Poisson distribution. Claim severity was modeled using a Gamma distribution. Estimates for overall losses were derived from the frequency and severity models.

Results:​ Forward collision avoidance systems, particularly those that can brake autonomously, along with adaptive headlights, showed the biggest claim reductions. Other systems, such as blind spot detection and park assist, did not show consistent effects on crash patterns across different manufacturers. Lane departure warning systems were associated with increased claim rates;​ however, the 95% confidence intervals were large, indicating the results are uncertain. Forward collision avoidance systems with autonomous braking showed 10-14 percent reductions in the frequency of claims to repair damage that the studied vehicles caused to other vehicles;​ adaptive headlights showed reductions of as much as 10 percent in the same types of claims. Consistent with this finding, injury liability claims also were reduced. Both systems were associated with more modest reductions in the frequency of claims to repair studied vehicles. Forward collision avoidance systems without autonomous braking also reduced claims for damage and injuries but to a lesser extent.

Conclusion:​ Insurance data show some collision avoidance technologies are preventing crashes and injuries. In the case of forward collision avoidance systems, the largest crash reductions involve damage to other vehicles. That is consistent with a reduction in rear-end crashes, the particular hazard these systems are meant to address. Adaptive headlights also appear to prevent collisions with other vehicles, but it is unclear why they aren’t more effective at preventing single-vehicle collisions. Insurance data provide a first look at the overall effectiveness of these systems, but detailed crash information is limited. Most systems can be deactivated by the driver, and the status of a feature at the time of the crash is not known. Deactivation could partially account for the lack of demonstrated benefit for certain features. These analyses estimate real-world effectiveness of several crash avoidance technologies. While some results indicate the need for further investigation, it is clear that certain systems, such as those that help drivers avoid collisions with the vehicle in front or better illuminate the road ahead, can play a role in making roads safer.