Events

AnimalTracker is a website that graphs GPS data from cattle in Eastern Oregon and Idaho, in order to help animal and rangeland scientists conduct their research. The main feature is an interactive webpage that visualizes GPS data logs collected from cow herds, for research and industrial purposes. This data then gets processed using a multitude of data cleaning and outlier detection algorithms, which can be configured by the scientist or rancher using the application, and turned into a variety of graphs and statistics about the herd's movements. These graphs and statistics allow rangeland scientists and cattle herders to better understand the movement behavior of the animals they are researching or working with. With this information, one can remotely study cattle behavior such as where they move to, when they move, and at what elevations. Version 1.0 of AnimalTracker was developed at Boise State University; over the course of our capstone program, our team has assisted in further developing the data cleaning algorithms and visualizations for Version 2.0. The base code and initial features were provided to us upon the start of the project by our project partners at Oregon State University and Boise State University, and we have expanded on these features in three main ways: implementing DBscan data clustering, Kalman data filtering, and by creating a time series animation to map the movements of the cattle across multiple days.
 

AnimalTracker can be found on the Comprehensive R Archive Network (CRAN). Being a part of CRAN means anyone can download and use Animal Tracker, and with the general accessibility of cheap GPS collars, this permits the application to be a cheap and effective option for those trying to track cattle herds for research or industrial purposes. Although this version of AnimalTracker is largely tuned for tracking cattle, it’s possible for it to be used in other circumstances as well, meaning that one could potentially use it for other herds of animals, for ecological research, or for visualizing other types of GPS data.

We'd like to thank our project partners for their leadership and technical implementation assistance:

• Sergio Arispe (lead researcher), Oregon State University
• Joe Champion (lead developer), Boise State University
• Thea Sukianto (student assistant), Boise State University