In this project, I demonstrate using an IMU sensor to perform reliable pedestrian dead reckoning (PDR) IMU sensor measures acceleration, rotation and magnetic field Walking dynamics produces noisy sensor data which is not useful for inertial navigation To overcome this, we can process sensor data into useful user motion based on a walking dynamics model
Assuming a constant device to user reference, we can infer user motion with user speed propagated by heading estimate User motion is best visualized in a trajectory plot shown below This data can be combined with GNSS receiver via sensor fusion to achieve high position accuracy in GNSS challenged environment This can also be used as standalone navigation in indoor environment