ANIMA:



Accelerometer Network Integrator for Mobile Animals: a New Instrument Package for Integrating Behavior, Physiology and Ecology of Wild Mammals. Much of our current understanding of the biology of terrestrial mammals is based on in situ observation, or GPS enabled smart tracking collars. Much of the impact of habitat encroachment and ecological changes to the environment is poorly understood, and the data is too coarse. A simple task such as determining what the animal ate is difficult to determine and will often require a search of likely places on the path marked by the collar.


Current wildlife tag technology runs “open loop,” with sampling set to occur periodically on the sensors, set once at deployment. We are designing a tag that can sense what the animal is doing (via signal processing techniques) and alter its sampling behavior accordingly (e.g., more frequently when hunting, less frequently when resting).


By utilizing modern low-power microcontrollers, and analyzing the data from sensitive and low-power three axis accelerometers and magnetometers, we are able to use signal processing techniques (such as Fourier and Wavelet analysis) to determine what “state” the animal is in. Based on this state, a new sampling schedule is determined for the sensors, as well as encoding the behavior tag into the low-bandwidth data stream.


Using a prototype collar deployed on a female mountain lion, we have already been able to determine when she was stalking her prey, her pounce (and miss), and her subsequent re-stalking and killing of her prey from the data stream. This also led to a metabolic analysis of the animal to determine how often she would need to kill to survive.


This project is making progress on creating a small “aware” collar to increase the fidelity of the data that the biologists have to work with. While the test deployment is on Mountain Lions, it has relevant capabilities for many other types of animals.



multimedia


  1. photos and images



Publications


  1. (1)Williams T., Wolfe L., Davis T., Kendall T., Richter B., Elkaim G., Wilmers C., “Energetics and mechanics of mountain lions: a step by step analysis for carnivore conservation,” Integrative and Comparative Biology, Vol. 52, Supplement 1, April 2012, pp. E193-E193, doi:10.1093/icb/ics078 (pdf)


People


  1. Gabriel Elkaim, Associate Professor, Computer Engineering, UCSC, 831.459.3054

  2. Maxwell Dunne, Masters Student, Computer Engineering, UCSC, 831.459.2140

ANIMA Project