Co-location Detection System Using Mobile Magnetometer Data

Co-location detection system is one of the most decisive strategies to stop or mitigate an epidemic as well as pandemic's outbreak. While many sensors, such as GPS, can be deployed to accomplish this task, magnetometer has emerged as one of the best solutions since not only does it have potential to achieve high accuracy, similar to that of GPS, but it also consumes considerably low battery power. In this paper, we proposed cloud computing and deep learning-based approach to perform co-location detection task from magnetometer readings. Experiment for this system, conducted in Busan, South Korea, has shown that our proposed idea can precisely differentiate whether two or more individuals were in the same location or not.