Apps and devices that track motion are of questionable use value – it’s true that they provide a general idea of a user’s activity level, but that can very wildly from reality, since there’s often very little involved in the motion tracking algorithms involved beyond detecting motion and counting that as a step. It’s not often the fault of these devices: as humans, we’re very inconsistent and unreliable in terms of where we keep these things.
Sometimes, your phone’s in your pocket. Sometime it’s in a bag. Sometimes it’s in a holster, if you miss your BlackBerry very much. But the point is, it’s not always in the same place, and where it’s stored can make a huge difference in terms of how accurately it gathers and reports information about the motion and activity of its owner.
A team at Northwestern University wanted to make activity tracking more accurate, without forcing humans to be more consistent in how they use and carry their devices. The result is an algorithm that not only compensates for when a device is carried in a sub-optimal location on a person’s body or in their bags, coats and purses, but one that actually predicts and detects when it’s being carried in different ways so that it can adapt to each new transport scenario without input from the user.
The study describing the NWU team’s work doesn’t deny that there are ways to carry a smartphone that detect motion more accurately than others – that’s still the case. But using the algorithm they’ve developed, people will at least be more likely to get an accurate picture of their daily activity regardless of where they store their phone. And the goal of the team isn’t just to make sure the next generation of NIke+ Move and ARGUS apps are better at telling you you’re lazy; the NWU researchers want to make apps useful in diagnosing and understanding critical diseases with motion-related symptoms, including Parkinson’s.
Study lead investigator Konrad Kording says that he believes smartphones will have crucial roles in personal health in the very near future, and part of preparing for that future involves making sure they not only have the sensors, but also the smarts to make sense of any data they’re gathering.