New algorithms can tell if you ride the bus, and how bad you drive

Researchers from the University of Helsinki in Finland have developed algorithms they say are the best yet for determining how mobile phone users are getting around town — ¬†walking, driving, taking the subway or otherwise — by analyzing the frequency and velocity of their starts and stops.

Paired with location data, this could be great news for civic planners trying to optimize roads or public transportation. As the researchers suggest, it could also be helpful in reconstructing accidents or identifying road hazards, or powering an app that gives feedback about a user’s driving style in the name of improving safety or fuel efficiency.

It’s arguable the latter capability is better suited for the car’s internal sensor network and display system, though. For starters, cars can gather much more fine-grained data about a vehicle’s operation than just braking and acceleration patterns. And should such a system turn into a glorified, digital backseat driver, a car-based system is much less likely to be thrown out the window.


Source: University of Helsinki