Abstract
Magnetic target tracking is a low-cost, portable, and passive method for
tracking materials wherein magnets are physically attached or embedded
without the need for line of sight. Traditional magnet tracking
techniques use optimization algorithms to determine the positions and
orientations of permanent magnets from magnetic field measurements.
However, such techniques are constrained by high latencies, primarily
due to the numerical calculation of the gradient. In this study, we
derive the analytic gradient for multiple-magnet tracking and show a
dramatic reduction in tracking latency. We design a physical system
comprising an array of magnetometers and one or more spherical magnets.
To validate the performance of our tracking algorithm, we compare the
magnet tracking estimates with state-of-the-art motion capture
measurements for each of four distinct magnet sizes. We find comparable
position and orientation errors to state-of-the-art magnet tracking, but
demonstrate increased maximum bandwidths of 336%, 525%, 635%, and
773% for the simultaneous tracking of 1, 2, 3, and 4 magnets,
respectively. We further show that it is possible to extend the analytic
gradient to account for disturbance fields, and we demonstrate the
simultaneous tracking of 1 to 4 magnets with disturbance compensation.
These findings extend the use of magnetic target tracking to high-speed,
real-time applications requiring the tracking of one or more targets
without the constraint of a fixed magnetometer array. This advancement
enables applications such as low-latency augmented and virtual reality
interaction, volitional or reflexive control of prostheses and
exoskeletons, and simplified multi-degree-of-freedom magnetic
levitation.