user-based insurance: great example of IoT’s pros & cons

Posted on 26th November 2012 in Internet of Things

Today’s NYTimes has a feature on the growing phenomenon of “user-based insurance,” in which your insurance rates are based on your actual driving behavior, not the proxies that insurance companies used in the past to make up for lack of information on your real driving behavior.

I was only aware of Progressive’s Snapshot, but the article reports that Allstate and StateFarm both offer similar services (I’ve read elsewhere that the concept is more widespread in Europe).

I think user-based insurance is a great example of the tradeoffs inherent in the Internet of Things.

First, the details: a small device clamps on to the diagnostic port on your car’s dashboard. Typically, the accelerometer then records data such as miles driven, time of day (the rate of accidents increases late at night), whether you make sudden stops, etc. (at least the Progressive one, and perhaps the others, does not include GPS, to avoid concerns about having your routes tracked). On the positive side, it can result in real savings for the driver, because the company can more confident that it can predict your likelihood of being in an accident than with some of the “proxy” indicators such as age that they’ve had to rely on in the past (my favorite: your credit rating. This avoids, as the story mentions, the crime of “driving while poor.”).

On the other hand, there’s the nagging concern about what a company, or, more likely, a rogue employee, could do with your information. Part of the concern is valid: Progressive mentions that the data could be subpoenaed in case of an accident. IMHO, this is one more reason why the US government must at least consider privacy and security protections for IoT users.

There’s already an important move afoot to maximize the benefits of the plug-in devices while protecting personal privacy. MIT’s CloudCar initiative is an attempt to create an open-source monitoring device whose users would be firmly in control of who could use their driving data and how.  As I wrote in my e-book on the IoT:

“One current telematics research project is noteworthy not only in its own right, but because it addresses one of the prime obstacles to the IoT: lack of global communication standards.

“The goal of the CloudCar project, initiated by MIT’s Field Intelligence Lab, is to create a universal, open standard that can be used throughout the automotive telematics industry. That’s important, because currently there is a bewildering array of communications protocols, since each of the car manufacturers and other vendors are working on proprietary projects, ranging from Ford and Microsoft’s Sync to GM’s OnStar.

“If any of the truly remarkable telematics applications, such as ones that automatically warn of impending road danger proposed by BMW on their ConnectedDrive platform, are to become a reality, a common standard is needed. As Erik J. Wilhelm of the Field Intelligence Lab says, ‘Manufacturers have no compulsion to share with other manufacturers, so there’s no opportunity for really great applications such as hyper-accurate traffic mapping which would only work if everyone is talking the same language.’

“MIT is currently self-funding the CloudCar project, which resembles Arduino in several important ways: the heart of the system is a simpler controller, now in it’s fourth generation, which would plug into the car’s diagnostic port, would be user-friendly, and would allow streaming of data from the car to a wide range of 3rd party vendors, who could use it for everything from  car sharing applications to automated roadside assistance systems.

“This hardware will be open-source, and the standard underpinning the data server will also be open. The Field Intelligence Lab hopes to have about 10,000 of the units in the field by the end of 2012. The users would pay a modest fee, with the remainder of the cost to be picked up by third parties that would be able to use the data the boxes yield if drivers grant them permission.  ‘Everyone needs the same data for all of these potential applications, although some need it more immediately while other need to make certain it is more accurate. Safety and security is central to our design efforts because we have to make certain cars won’t just stop in the middle of the road,’ he says.  If the CloudCar standard is widely adopted, application developers could be gathering and sharing information from millions and millions of cars.

“The mindset fostered by the CloudCar project could be as valuable as the standard itself.

“ ‘Ford and GM both have announced they were opening up their proprietary products such as OnStar, but they are  only giving access to big companies, not the creative commons space. If you’re all speaking the same language, everyone wins and it’s still possible to have a survival-of-the-fittest application market that mobile phone users have come to expect, only this time with many more data streams and possibilities,’ Wilhelm says.

“Wilhelm hopes that the CloudCar standard will have value beyond telematics, since other crucial segments of the IoT, such as medical reporting and home applications are also handicapped by the lack of communication standards. ‘CloudCars are nearer term, offers the most immediate gains, and faster adoption rates than these fields,” he says. “We hope to further apply our learning to CloudHome and CloudMe.’”

Aside: why don’t companies promote user-based insurance more aggressively? Progressive does run ads for it (although I don’t think most people would understand the  concept from the ad), but I hadn’t seen any promotion for the Allstate and StateFarm services.

comments: Closed">Stephenson blogs on Internet of Things Internet of Things strategy, breakthroughs and management