Telematics for insurance: Analytic insight is a many-splendored Internet of Things

Senior Managing Consultant, IBM

In a recent post, I asked if a vehicle telematics use case was overstated. From the tone and substance of the blog it appeared that I was driving at a definitive “yes”—but then I changed direction and suggested that it was not overstated but missing input.

To be clear, the use case I was questioning is where vehicle telematics and insurance analytics, by virtue of an installed dongle or smartphone app used to capture driving behaviors, can actually become predictive of loss. My answer is supported by a number of recent articles (such as Celent, LexisNexis and so on) that say no (or at least, not yet).

So, here’s what I mean.

I think the industry has gotten lost in the blinding light of all the new data that is suddenly available. Vehicle telematics is no exception. With telematics for insurance, the goal of underwriting an auto policy based on each driver’s behavior appears to be within the industry’s grasp. But not so fast. Sure, the data being captured provides new views and delivers more expediency in volume compared to previous industry capacities. But is it enough? Is it accurate? What don’t we know? my earlier blog I offered this explanation:

“This improvement in the technology, using data from a telematics sensing device is thought to provide significant insight into a driver’s behavior. But wait, if that were true, why haven’t insurance companies leveraged this data to create more accurate pricing plans? There are a couple of issues:

  • The industry determines their pricing by driver. In a multiple driver family, telematics cannot determine who is driving (yet).
  • The driver, whoever they may be, can simply unplug the dongle or turn off the ‘app’ in order to suppress the data collected. Essentially, posting incomplete data.
  • While some of the context around driving behavior can be analyzed to form insight, not all of it can be captured. It is that contextual data that offers the greatest opportunity to understand the behavior of the driver.”

I came across a September 2015 Fortune magazine article that was calling out some of the new capabilities out there. Many are pretty cool and the insurance industry is paying attention. Here is an overview of a few developments on the horizon:

  • High Mobility: A German firm that may answer the question “who is driving”
  • Trilumina: Development of semiconductor lasers that “see” as part of the self-driving car
  • Driversiti: An app in your phone learns driving behavior and alerts you if there is a change in performance, such as sloppy steering (distracted, sleepy, impaired)
  • Sober Steering: A touch-based sensor that can determine blood alcohol level

My favorite, however, is Nebula Systems—a cloud-based platform that claims to aggregate all the sensor data that cars generate today, which I believe offers a greater level of context from the data and therefore, more meaningful insight. And that is the point of this blog entry. The Internet of Things (IoT) is providing us with more data than ever before, but to make better sense of it—that is, to get to those crisp insights and useful applications where the industry can find value—requires multiple “things” from the Internet of Things.

My next blog will discuss a creative integration of multiple IoT sources and show how this is turning heads in the P&C industry, because analytics insight is a many-splendored Internet of Things. Learn more about the different use cases for insurance business intelligence via telematics, data and analytics, and request a complementary workshop tailored for your business.