GRAZETOOLS: A Set of Tools for Analyzing Livestock Behavior Using GPS data

Abstract:

As Internet of Things (IoT) devices (e.g., GPS, accelerometers, magnetometers, gyroscopes) become ubiquitous, they are increasingly relied upon to collect data and to monitor assets for geospatially explicit scientific research. Analysis of animal-wearable sensor data to characterize livestock movement patterns and to inform grazing management decisions is an example of current IoT applications in agriculture. The generalized use of IoT devices leads to the rapid growth of data that are collected through a variety of sensors. Analysis of large volumes of data without automation capability is laborious, time-consuming, and can lead to erroneous or inconsistent answers to scientific questions. Domain researchers, often lack sufficient programming and computing background, and therefore face difficulties in processing massive amounts of sensor data. We developed GRAZETOOLS: a set of tools to support the integrated analysis of cattle behavior on rangeland by analyzing data collected by GPS devices. GRAZETOOLS consists of three tools: 1) GRAZECLASS to partition a GPS data collection according to given criteria, such as the time of sunrise and sunset, 2) GRAZEPIX that tracks pixel use from GPS data of grazing pastures, such as percentage of grazed pixels, the re-visitation rate of each pixel, etc. 3) GRAZEACT that calculates activity parameters from GPS data of grazing animals including distance traveled, path sinuosity, vegetation preference index, convex hull points, convex hull area, etc. We will demonstrate the use of the tool set by using several GPS datasets with cattle grazing locations and show how the tools can help automate calculations in studying livestock behavior and managing grazing on rangeland.