Human Gait Based Relative Foot Sensing for Personal Navigation

Abstract

Human gait dynamics were studied to aid the design of a robust personal navigation and tracking system for First Responders traversing a variety of GPS-denied environments. IMU packages comprised of accelerometers, gyroscopes, and magnetometer are positioned on each ankle. Difficulties in eliminating drift over time make inertial systems inaccurate. A novel concept for measuring relative foot distance via a network of RF Phase Modulation sensors is introduced to augment the accuracy of inertial systems. The relative foot sensor should be capable of accurately measuring distances between each node, allowing for the geometric derivation of a drift-free heading and distance. A simulation to design and verify the algorithms was developed for five subjects in different gait modes using gait data from a VICON motion capture system as input. These algorithms were used to predict the distance traveled up to 75 feet, with resulting errors on the order of one percent.