The 5S is a game changer for location & motion centric apps.
With the new iPhone 5S apple is clearly pushing for more and better location based/lifelogging apps like Rove, Foursquare or Moves.
These apps run in the background and use GPS and accelerometer sensors to automatically guess where you are and what you are up to. The precision and frequency of their measurements is somehow correlated to the battery drain and the poor quality of GPS signals indoors. This is about to change with the 5S. Here is why…
1 — The M7 co-processor
On IOS6, apps like Rove or Foursquare periodically measure accelerometer and/or GPS data to detect your general location and to guess whether you are stationary/walking/driving/flying and more. In the case of Rove these measurements are made every couple of minutes to minimize battery consumption, and even less frequently if you are stationary or moving fast. This approach is not always possible for apps that need to measure every step you take or every move you make like most fitness apps. Thanks to the M7, access to continuous motion data now comes at no additional energy cost!
Motion logging apps will no longer need to run nonstop in the background and instead will be able to use the data collected by the M7. This means that the overall energy footprint will drop, at the same time the precision and quality of your data will improve.
Some might wonder why adding an additional processor that continuously tracks your motion would increases the overall lifetime of a full battery charge. The M7 is a small processor optimized for one task (measuring your movement) and therefore has a much lower energy footprint than the main processor for the same task. It will also allow the main process to be dormant more often and only “wake-up” when a motion is detected, similar to how your screen saver works. For example the operating system could reduce the ping frequency for new networks if you are not moving.
2 - The new motion API
With the 5S, developers will not only have access to raw vectorial data but also human readable motion states : walking, automotive, running and stationary. At Rove this is something that we currently have to do in-house. We developed models using raw acceleration data to guess your motion states. This new API won’t replace our current models but will allow us to improve them and focus on other aspects of our “guessing algorithm” for example knowing if you are riding a boat or a train vs just “an automotive” transport, similarly to how we currently detect flights and more. With this new API Apple is clearly encouraging developers with no data-mining background to build new pedometers, fitness apps or other types of motion aware apps..
3 - iBeacon
One of the major problems, location based apps run into, is the accuracy of indoor GPS measurements. GPS was designed as an outdoor localization system. Satellite signals are often blocked by building structures or bounce on walls in cities like New York resulting in a very imprecise read on your location. This is sometimes improved with wifi signals or cell towers triangulation but not nearly enough. Obviously this reduces apps ability to detect the exact venue you’re in. To mitigate for this issue Rove uses your entry and exit points to approximate indoor locations. This is about to get much easier and much better thanks to iBeacon. It will be as if there was a couple of satellites in every single venue… apps will now get your coordinates in HD meaning they will not only know which restaurant you’re in but also which table you’re sitting at, creepy? maybe a little… . That’s the kind of precision iBeacon will enable! This will improve the quality of existing location based apps and open the door to a whole new set of features, for example keeping track of which paintings you stared at the most in a museum…imagine that!
