Abstract
This paper presents a wearable system for measurement and monitoring human body joint angles based on UWB ranging. The DW1000 chip was used with standard deviation of distance measurement within 10 cm with range up to 70 m. We propose a method for enhancing range measurement accuracy based on an estimator which compensates clock imperfections and relative pairwise movement of nodes. Since the estimator is valid only for small slices of time, we propose continuous motion estimation algorithm based on segment-by-segment data processing and stitching results into a final solution. The pairwise distances are approximated with Taylor series of a given order L in short measurement windows while timestamps are compensated with clock parameters of a first-order clock model. The main contribution of the proposed method is the ability to implement joint angle estimation by using low-cost off-the-shelf UWB components, without high-precision clock sources or a need for wired or wireless time synchronization. In order to determine an optimum order L and time slice length, Sprague and Geers' metric was used. The method was experimentally evaluated in static and dynamic conditions. The results show that the accuracy of the proposed system is comparable to similar solutions based on laboratory equipment.
Highlights
Human motion tracking (HMT) has gained a lot of attention in the last decade
In this paper a wearable measurement system for static and dynamic monitoring of human body joint angles based on UWB ranging approach was developed and tested
Sensor nodes are based on the DW1000 integrated circuit, which provides timestamps with 15.6 ps accuracy, what translates in approximately 10 cm of standard deviation of ranging error
Summary
Human motion tracking (HMT) has gained a lot of attention in the last decade. Studying kinematics, or more motion of human body, is of interest in various areas of research. Solutions without markers, which are less intrusive and easier to setup, such as Leap Motion, do exist, but they are either limited in the field of view or in their estimation accuracy Their robustness is inferior in comparison with conventional marker-based computer vision systems [2]. There are alternative systems based on pure mechanical solutions that use potentiometers and pulleys attached to tracked limb so that joint angles can be determined [5] Those kind of systems are accurate but suffer from alignment issues which cause lower dynamic response. Due to absolute accuracy localization, UWB systems are immune to the most serious IMU sensors drawback such as accumulated drift error They do not suffer from occlusion problem since radio signals can penetrate human body and most of occlusions. Our solution was validated on JACO Robotic Arm in various scenarios
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
