Breakthrough

Multisensor Fusion

  • Wearable Inertial Measurement Unit (IMU) sensors are being widely deployed in various fields where good estimates of the orientation of a rigid body are required. Examples include orientation estimation of the human body or a particular segment for various activity related and healthcare-related applications, interactive gaming and robot prosthetic body parts orientation estimation to provide stability and control. Our work in this area has resulted in the following journal publication in IEEE Instrumentation & Measurement Transactions

GNSS Receiver Design

  • Global Navigation Satellite System (GNSS) receiver provides position, velocity and timing estimates anywhere on the earth by measuring the distances from at least four satellites. The accuracy of these estimates highly depends on the distance measurement. We have been working on different ways to improve the accuracy of these distance estimates which in turn could improve the accuracy of estimated position, velocity and time. The work has produced the following journal publication in IEEE Communication Letter:

    In the paper, we have proposed a novel approach to improve the position estimates based on this idea. The proposed approach is based on extracting the principle components from singular spectrum of the code delay measurements inside the receiver. 

A. Hamad and M. Tahir, "Improving the Accuracy of Human Body Orientation Estimation with Wearable IMU Sensors" accepted for publication in IEEE Transactions on Instrumentation & Measurement.

  • In the paper, we have proposed a novel Kalman filter to accurately estimate the orientation of different human body segments wearing IMU sensors. The data from IMU sensors is fused inside the proposed Kalman filter to get the orientation estimates. The presence of linear accelerations in accelerometer measurements causes significant problem during sensor fusion process. The proposed approach estimates the direction of these linear acceleration and assigns lower weights inside the Kalman filter to only those sensor axes which are experiencing these acceleration. In this way, we effectively remove the problem of linear acceleration which results in good orientation estimates.