Scientists at Rutgers university have developed a method of only by the speed of a car to measure, to determine the route a person has taken. This what they want to show that also with only this data, the privacy can be.
The method, which the researchers elastic pathing to call it, can be a exact route a driver has to reproduce without the need of gps or similar techniques can be used. This is only a starting point and an obd connector in the car, which the speed via the vehicle management system can be read out, and via bluetooth can be forwarded to, for example, a smartphone.
With the software you can on the basis of the speed where a driver in a car to go driving. The algorithm looks at how fast someone is driving, and compare that with the roads on a map from the starting position. As soon as the speed changes and the image no longer corresponds to the assumed way, the software searches for roads near where that is the case.
The scientists conducted with six directors 240 rides in New Jersey with 46 destinations. In addition, they held a similar test with 21 directors, of which 691 rides with 240 destinations reason in Seattle. From those experiments turned out to be the method for 26 percent of the destination is correct to determine, with a margin of error of almost half a kilometer.
Absolutely perfect, the method is not, emphasize the scientists. Drivers differ from each other in terms of driving styles and in addition must be taken into account the situation on the road at that time. Nevertheless feel they can show that the location of the driver can be traced, while, for example, gps is used.
With that observation, assign the computer scientists to American insurance companies, that the speed of cars more often obtain to the behavior of their customers. “The data, once collected, no more road. Improvements make it possible for more personal information to pick up”, as they say. They find that alternatives should be considered, to the privacy protection. During the International Joint Conference on Pervasive and Ubiquitous Computing following month to present to the scientists, their final findings.
Comments
(84)