advertisement
advertisement
advertisement

What 1 Billion Miles Of Driver Data Tells Us About How We Drive

Insurance companies are using data to try to shape driver behavior and reduce accidents–here’s what they’ve found so far.

What 1 Billion Miles Of Driver Data Tells Us About How We Drive
[Photo: Flickr user David Pacey]

While cruising down the highway may feel to a driver like a solitary activity, vehicles are more connected than ever. Hundreds of millions of drivers worldwide use services like Waze that log how fast you’re driving and where you stop. If you live in the United States and use an insurance service like Progressive or Metromile, the odds are good you have a little device that plugs into your dashboard that logs and shares information about how much gas you guzzle, when your car is stuck in traffic, and how your car’s insides are holding up.

advertisement
advertisement

Now LexisNexis, one of the world’s biggest data aggregators (and a company that makes hefty profits working with insurance companies) is examining a collective one billion miles of data driving and has identified some major trends: Female drivers are better drivers than males, you can identify which individuals are most likely to speed, and hard accelerating, braking, nighttime driving, and speeding are the main precursors to traffic accidents.

“We looked at one billion miles of driving, and the claims that go with it,” LexisNexis’s senior vice president of auto and home insurance Ash Hassib told Fast Company. “You could see speed events, time of day driving, and hard braking events. The goal of this is for companies to try to match claim history with driving data. If you drive like this, you’re twice as likely to file a claim.”

Last year, LexisNexis Risk Solutions acquired a company named Wunelli that helps insurance companies mine data from automobiles through smartphones and dashboard devices. LexisNexis and Wunelli work with insurers around the globe through the telematics integration process, and also manage the data they collect. The company began logging mile-by-mile smart-car information beginning in approximately 2008. In March, they reached the aggregated one billion miles of data collected from their insurance clients’ customers.

Hassib expects this sort of logging to become commonplace in the insurance agency over the next decade. A big part of it, he says, is the use of in-car smartphones and the push by automakers to integrate 4G antennas into their newest models. While plug-in telematics devices by insurers like Progressive and Metromile (who aren’t LexisNexis clients) and “Fitbit for car” products like Mojio require buying a separate product and can sometimes be a hassle to plug in, iPhones and Androids are currently much more useful for insurance companies. While their sensors aren’t as sophisticated, they’re already in the car and already have robust accelerometers, GPS tagging, and other functionality… in short, it’s a lot cheaper for insurance R&D divisions to use smartphones instead of plug-in devices.

Insurance companies are tracking your every turn both to save money and build better risk models, Hassib told me. “Motor vehicle records only show when you’re caught,” but telematics devices and smartphone apps are a “continuous monitoring tool that assesses, nonstop, how drivers are doing.” In order to build usable risk models, he added, they need data from at least 30,000-40,000 “earned car years,” a measure of risk used by insurance companies. One earned car year approximately equals one year of driving on an insured car. Otherwise, the data isn’t reliable and the sample size is too small.

One product Hassib and LexisNexis are particularly proud of was made for an Australian client, the insurance firm AAMI. AAMI released a Safe Driver App that uses sensors inside Androids and iPhones to track aggregated data about driving speed, braking habits, acceleration style, fatigue, and–crucially–whether drivers use Bluetooth in transit to use their phone, or play with the handset as they drive. Drivers–who consent to turn over the data when they use the app–then get to see their score rated on a scale of 0-100 (with high-scoring drivers being told they’re “Smoother than Barry White”).

advertisement

The data LexisNexis and their competitors collect isn’t perfect. One challenge Hassib’s team deals with frequently is cleaning data from Android phones–the process to cleanse data from a Samsung phone is much different than that of a LG, for instance. Cell tower triangulation is also an issue: Weaker signals in rural areas mean less quality information. For insurers, having dense 4G networks is a necessity.

For drivers who want to avoid being data-mined, the answers are twofold: Use an old-fashioned map or written directions instead of a smartphone app for navigation, avoid telematics plug-ins, and drive an older used car with less digital integration. Meanwhile, both auto companies and insurers are planning on making attractive offers designed to get drivers to agree to tracking: Connected cars have goodies like Pandora on the dashboard or Netflix in the backseat for the kids, and insurers will offer steep discounts in exchange for access to car telematics.

Correction: An earlier version of this article incorrectly stated the date of Wunelli’s acquisition. It was in 2014.

advertisement
advertisement