Robust human motion tracking using wireless and inertial sensors

Recently, miniature inertial measurement units (IMUs) have been deployed as wearable devices to monitor human motion in an ambulatory fashion. This thesis presents a robust human motion tracking algorithm using the IMU and radio-based wireless sensors, such as the Bluetooth Low Energy (BLE) and ultra-wideband (UWB). First, a novel indoor localization method using the BLE and IMU is proposed. The BLE trilateration residue is deployed to adaptively weight the estimates from these sensor modalities. Second, a robust sensor fusion algorithm is developed to accurately track the location and capture the lower body motion by integrating the estimates from the UWB system and IMUs, but also taking advantage of the estimated height and velocity obtained from an aiding lower body biomechanical model. The experimental results show that the proposed algorithms can maintain high accuracy for tracking the location of a sensor/subject in the presence of the BLE/UWB outliers and signal outages.