Enabling energy-efficient driving route detection using built-in smartphone barometer sensor

Human mobility is known to follow simple reproducible patterns, i.e., humans tend to travel a few known places. Early detection of those “significant journeys” has a prospect for emerging smart applications like real-time traffic route recommendation and automated HVAC (heating, ventilating, and air conditioning) systems. In this paper, we design, implement, and evaluate a mobile system that effectively captures significant journeys based solely on the embedded barometer sensor of a smartphone. Unique elevation signatures of driving routes are extracted using the barometer sensor of a smartphone that is position/orientation independent and exceptionally energy efficient compared to other power-hungry sensors like gyroscopes and GPS units. Using over 90-miles of real-world data collected from the city of Brookings, SD, we demonstrate that the proposed system achieves the mean detection accuracy of 95.7% with the mean false positive rate of 2.3% using only the barometer sensor.