Every month they send me an email with my km covered by each mean of transport. But how does the algorithm tell if i was riding my bike, or in a slow car? How can it distinguish buses from car? Only because buses stop more often, or because there’s more people with Android devices on the bus?
Thanks!
In: 3
I hope a Google engineer chimes in, but my understanding is they compare a few metrics to determine your mode of transportation:
-speed you’re traveling at
-rate of acceleration
-which paths/roads you’re traveling on
-are you generally following “automobile rules” like stop lights, one way streets, etc
-are you generally following “public transportation rules” like following certain predefined routes, stopping often, etc
Between these metrics I bet a computer could predict how you’re traveling 99/100 times!
The speed you travelled at (using your GPS to work it out) and the points at which you stop.
Google may know that you are in a car by your Bluetooth connection.
Google may know that you are on a bus if you are in close proximity to a dozen other people.
It may know you are traveling by bike by the motion on your legs peddling or the fact you’re going slower than a bus and you are by yourself.
Google also figures out traffic the same way. Many phones all going going slow.
Not sure how they do it exactly but a few commenters have provided some insight.
Add to this, Google doesn’t only track your GPS. It’s also harvesting the travel data of everyone else. If you’re on a bus, your GPS will be extremely similar to those of at least a few other people for a period of time. Google also has bus schedules / GPS for many cities to cross-reference. In my city there are transit apps that already take data from riders to improve accuracy of real time bus schedules.
If you’re in a car vs bike, the speed, acceleration, when you stop vs when other people stop for traffic lights etc will generally give them an idea. If you go on bike paths or trails, that would be even more telling.
Latest Answers