First of all, we should thank Pranav Pasricha
and support the Swiss Re initiatives. Working on, as we think, the major problem in commercial insurance, we stand by every word of Pranav:“The variety of telematics data sources and types is the main obstacle for introducing telematics-based service on a scale.”
After countless attempts and failures to introduce telematics in insurance, we ended up with a toxic image of telematics in the industry and a highly fragmented data landscape. And this data is as diverse, as it’s difficult to get, understand, and work with.
Speaking about the diversity of data sources, we agree on the problem of data quality and resolution, i.e., how rich the data sets are and how frequently the information is updated.
For example, mobile apps fail to provide rich data, unlike OEMs that generate hundreds of parameters. At the same time, aftermarket telematics providers offer fewer parameters with higher frequency or resolution.
Poor resolution and frequency affect risk assessment, and we should either learn how to enrich it or wait 15 years till OEM embedded data is on the scale and affordable.The issue of telematics data implementation
One more aspect to consider is how insurers use telematics data for risk assessment.
Traditionally, people associated telematics insurance with UBI, or mileage-based insurance, simply because the information about mileage is easy to extract and utilize. That’s why UBI has been on the agenda for years.
When it became clear that mileage alone is not equal to vehicle risk, telematics insurance scaled to behaviour-based models, that take speedings, brakings, and accelerations into account.
Still, one more data set is often neglected, although being an integral part of the game. And this is the contextual information – geography, road types, weather, and other conditions.
Only when we understand how the driver uses the vehicle, performs at the wheel, and under what conditions, we can build an effective risk assessment model. And this is, we believe, is the only efficient way to use telematics in this context.