> Drawing from existing literature, we found that accelerometer data alone may be sufficient to obtain information about a device holder's location, activities, health condition, body features, gender, age, personality traits, and emotional state. Acceleration signals can even be used to uniquely identify a person based on biometric movement patterns and to reconstruct sequences of text entered into a device, including passwords.
One I could think of is you could use accelerometer data to know if the phone is moving, and how: is it just being lifted and interacted with, is the user walking or running, are they in a car.
For a high profile target you might also be able to track their approximate location if you know their starting point and their acceleration profile - speed up, slow down, turns, time taken. Enough, at least, to execute an ambush.
With enough data and a high SNR, you don't even need the starting point or the velocity profile. Just the turns and distances are enough in combination with a map to uniquely identify most journeys. The problem with applying that more globally is sensor drift.
Because it can be used to roughly track location through 3D space. It's nowhere near as accurate as practically any other method but for "good enough" a lot of the time it'll do the job.
