At DIA Prime Time Swiss Re’s Automotive and Mobility Solutions team presented their ADAS (Advanced Driver Assistance System) Risk Score – and how this supports insurers, mobility providers and OEMS. Now, 6 months later, we sat down with Luigi Di Lillo, who leads Products and Partnerships for Swiss Re’s Automotive and Mobility Solutions, to talk about the latest developments in the ADAS Risk Score and the impact on insurance claims.
The first version of the ADAS Risk Score was developed and released in 2019, in collaboration with the BMW Group. Since then, we've continued to invest in the concept, adding new features and capabilities as new technologies become available. The initial idea was quite simple: we wanted to translate the impact of car safety technologies into insurance terms. Our goal was to generate higher adoption rates of safety technologies by broadening users' knowledge of how safety systems work – which in turn should contribute to better road safety.
Our second major breakthrough came in September 2020 with Toyota Insurance Services joining the ADAS Risk Score platform. As part of this partnership, Toyota and Lexus vehicle data has been integrated into our insurance scoring. Moreover, Swiss Re has recently tested and validated the development of its non-VIN-based scoring system, which is available worldwide for all vehicle brands.
The ADAS Risk Score has received positive feedback from insurers in the EU and APAC regions due to the added value that this new risk factor has delivered to their customers.
It's difficult to summarise the quantitative impact of ADAS features on road safety and insurance claims in one or two numbers. It’s a complex question, but broadly speaking, there are two main factors we look at:
First, the performance of the safety technology fitted on the vehicle: different generations of the same ADAS feature can deliver varying results even within a specific car brand. Figure 1 illustrates the reported difference we've observed in the impact that two generations of the same ADAS feature have had on the insurance risk premium. They were sold by the same manufacturer, under the same commercial name, and priced in the same way by the insurer in a European country. The ADAS feature assessed was an AEB system that recognises pedestrians, including stationary ones, and automatically triggers the vehicle's brakes to prevent or reduce impact. The first generation of this ADAS feature correlated with a 10% reduction to the technical risk premium, while the second generation proved to be much more effective, correlated with nearly a 40% reduction of the technical risk premium. This improvement stems from the fact that the newer version of the AEB is based on a more advanced camera technology and activation logic.
Active in a wider speed range, it combines an internal logic leading to sharper braking and faster speed reduction in an emergency. Such granular assessment is only possible thanks to the sharing of knowledge within the scope of Swiss Re's strategic partnerships with key players in the automotive industry.
Second, the traffic dynamics to which the technology and its functions are exposed: despite some slight differences across markets, on average we have observed that ADAS features have a consistent impact on safety and generally lead to fewer claims. Figure 2 indicates the reduction in claims reported as a result of three different categories of ADAS features in the United Kingdom and Germany. In addition to claims being reported less frequently in each country, we also observed a slight variation of the features' impact within each country (not represented in the figure), depending on the client's insurance portfolio exposure and location. For instance, a parking system will have a bigger impact on reducing claims in highly concentrated urban areas than in rural areas.
Swiss Re therefore typically recommends ADAS Risks Scores to be calibrated to the specific portfolio exposure for optimal results.
Absolutely. Traditional insurance risk models tend to overestimate the frequency of claims for vehicles that are less prone to accidents and underestimate the frequency for vehicles that are more prone to accidents.
The Swiss Re ADAS Risk Score can correct this bias. Analyses of European insurers' portfolios indicate that the predictive power of the motor risk model can be significantly boosted when using the ADAS Risk Score on top of the traditional base model. For MTPL (Motor Third Party Liability) coverage, such improvements include changing the technical rates by up to 30% for certain risk segments of the insurers' portfolios, for about 60% of the policyholders in each portfolio tested. The predictive power of the ADAS Risk Score in terms of MOD (Motor Own Damage) is equivalent to that of MTPL, with an increase of up to 30% compared to traditional technical rating models for over 50% of the policyholders. Contrary to widespread assumption in the industry, the portfolios tested so far indicate that the impact of ADAS on MOD cover does not cancel out the positive effect of ADAS on MPTL cover.
This shows that reflecting the impact of ADAS in insurance terms can be a remarkable predictor of risk for the motor insurance industry, which in turn can be used to develop strategies to allow for increased policy volumes and reductions in loss ratio.
Sure. Generally speaking, we have two options: the most accurate way of identifying a vehicle is through its VIN, a 17-digit code that is unique to the vehicle. ADAS Risk Score uses the vehicle's VIN to connect to granular vehicle build data, namely very detailed information regarding the vehicle equipment, including the technical specifications for the wide range of features and options installed. However, we know that VINs are not always collected during the insurance quoting process – even when they are, insurers rely on car registration numbers to find the VIN, which incurs costs.
For this reason, Swiss Re uses also a different, more flexible approach to vehicle identification based on a combination of distinctive car attributes, like the make, model, version etc. We help clients during the intricate pairing process that matches the vehicle attributes to our internal vehicle knowledge. The downside is that, without a vehicle's VIN, it can be difficult to take all of the optional equipment bought by a customer into account, which might lead to lower predictive power compared to VIN-based ADAS Risk Scores. However, as mentioned above, a vehicle identification process not based on the VIN may be the only option for some insurers. At Swiss Re, we believe the two approaches are complementary.
Figure 3 indicates that some car brands include most ADAS features as standard across their models, especially those relating to safety. Other brands have a different marketing approach. They offer customers a wide range of optional ADAS features that can be installed in their vehicles.
Luigi Di Lillo is leading the Products and Partnerships team at Swiss Re -Business Unit Reinsurance. He is responsible for spearheading solutions around vehicle safety, electrification, connected cars and vehicles with increased degree of automation. Luigi is a steering committee member of the World Economic Forum Safe Drive Initiative aimed at creating a generalized framework around the assessment of autonomous vehicles and fostering their adoption. He has been at the core of the recent Swiss Re partnerships announcements with Veoneer and Toyota.