Automated kinematic tracking using inertial sensor arrays


Meeting Abstract

P3-163  Monday, Jan. 6  Automated kinematic tracking using inertial sensor arrays SENTHIVASAN, S*; ALTSHULER, DL; University of British Columbia; University of British Columbia shreeram@zoology.ubc.ca

Recording kinematics with high precision and time resolution is a problem common to nearly all studies of animal locomotion. However, current optical methods depend on expensive high-speed camera arrays, are difficult to set up outside of laboratory conditions, and often have limited effective measurement volumes due to occlusion of marked points. Recent advances in microelectronics and the rising popularity of small-batch circuit assembly provide a cheap and accessible way of designing project-specific animal-mountable data loggers on the order of 2-4 grams. Further, Monte Carlo state space modelling methods provide a robust way of inferring kinematics from raw magnetic, angular rate, and gravity (MARG) sensor data. In the present work, we use one such logger to track the flight kinematics of pigeons (Columba livia) flown in flight chambers. To test the accuracy of this method, we compare the results with tracking data collected from an optical motion capture system. Animal-mounted MARG sensor arrays could provide a minimally invasive and automated means of recording animal kinematics at sampling rates comparable to that achieved by high-speed video. The low cost and portability may also facilitate field studies of animal locomotion.

the Society for
Integrative &
Comparative
Biology