The effects of driver assistance and automated driving systems on road safety and accident loss costs are the subject of numerous and often speculative prognoses. For some, a massive reduction of road accidents and attendant loss costs is more or less around the corner. This view is typically supported by the accurate observation that (depending on the market) about 90% of serious accidents are caused by human error.

However, this simplistic view disregards potential currently unknown or little-known causes of accidents. Many continue to stress the risks associated with the new technology and question the general superiority of the machine over the human driver in the context of dynamic and complex road traffic. Another hugely important aspect to consider is the adoption rate and speed of the most recent technology.

Against this background, the German Insurance Association, GDV, has formed an interdisciplinary project group comprising engineers, mathematicians, insurance experts and accident researchers. The group was charged with developing a realistic and well-founded picture of the effects that one can expect to see from assisted and automated driving.

The results of this in-depth analysis by these experts and their prediction of the impact on loss cost development up to 2035 was recently published in the paper “Automated Driving - Impact on Loss Cost Until 2035”.¹

The conclusions of the analysis show that modern driver assistance systems and automated driving systems will:

1. Make driving safer, although in practice they will prevent significantly fewer accidents than in theory.
2. Gain ground only with considerable delay, and therefore reduce loss costs only slowly.
3. Lead to higher repair costs in the event of damage.
4. Have a greater impact on MTPL (motor third party liability) insurance than on MOD (motor own damage) insurance.

Overall, the analysis shows two opposite effects. On one hand, fewer accidents will lead to lower compensation payments; on the other hand, the use of sensitive technology will increase repair costs.

Expressed in figures, the expert group concludes that driver assistance systems and automated driving systems in passenger cars can reduce loss costs by 7% to 16% by 2035 compared with the 2015 reference year. This overall result is made up of a reduction of 10% to 21.5% for MTPL and of 3% to 7.5% for MOD.² The final result will depend on the adoption rate of driver assistance systems and automated driving systems over time.³

In order to determine the outer points of the range, the project group selected two driver assistance technologies with very different adoption patterns. ABS (anti-lock braking system or anti-skid braking system) was introduced in the early 1970s, and it took the technology 20 years to be present in 40% of all vehicles on the roads. ESC (electronic stability control), also called ESP (electronic stability program), introduced from 1995 onwards, was adopted much faster, i.e. in 80% of all vehicles after 20 years.

While this modeling exercise based on past experience can clearly be viewed as plausible, it is perhaps not outside the realm of possibility that a future paradigm shift or a combination of individually less significant factors could speed up the adoption rate.

For instance, a sharp increase in shared driving or more kilometers/miles driven by highly automated vehicles might influence such a development. If that were to occur, the projections would only be valid for a period ending before 2035, and the corridor for reductions until that date would move upwards.

Endnotes

1. GDV-Themenschwerpunkt: Automatisiertes Fahren - Auswirkungen auf den Schadenaufwand bis 2035, https://www.gdv.de/resource/blob/8282/c3877649604eaf9ac4483464abf5305d/download-der-studie-data.pdf.
2. The Management summary presents the numbers for all vehicles including trucks and buses (7% to 15% for motor combined, 9% to 20% for MTPL and 3% to 7% for MOD).
3. The project group analyzed six areas of technology: four advanced driver assistance systems (park assist, emergency brake assist, lane change assist, and lane keeping assist) and two automated driving systems (motorway pilot and highway & city pilot, both supported by adaptive cruise control). The individual findings in each area were subsequently combined.

About the author

Mathias N. Schubert is an attorney-at-law and has been working in casualty reinsurance for 30 years, predominantly in Europe and also the United States. He now practices law with a focus on liability and insurance law, and provides consultancy services in the area of liability insurance and reinsurance.