Machine Learning-Driven Analysis of Smartphone IMU Data for Position and Step Counting

Smartphones have become extremely useful instruments in the field of ubiquitous computing for tracking and interpreting user whereabouts and motions. Modern smartphones are equipped with a wide array of sensors, which results in the generation of a plethora of data reflective of various positional states. This study investigates how smartphone accelerometers in particular, can be used as Inertial Measurement Unit (IMU) sensors to detect human steps and determine the position of the device. We can also accurately determine the phone's orientation by carefully examining these signals. In addition, we process accelerometer data using machine learning methods in order to detect and count steps and localization of phones which is then deployed on the smartphone. This methodology delivers important information about user behavior and phone location by utilizing the advanced sensor suite and broad availability of smartphones.