Date of Award

2016-01-01

Degree Name

Master of Science

Department

Computer Engineering

Advisor(s)

Eric W. MacDonald

Abstract

Global beef production is projected to reach approximately 60 million tons in 2015 and opportunities to increase efficiencies are significant with recent advancements in remote sensing. By economically monitoring the location and conditions of a herd such as feeding patterns, body temperature, heart rate, and other biometric information, the herd performance can now be enhanced through data-driven optimization. Next generation sensors provide high fidelity data on a wide range of measurable properties, and simultaneously, the wide adoption of these sensor systems in many consumer applications is leading to commoditization and reduced cost due to the associated economies of scale. Deploying sensors in feedlots requires overcoming many challenges including harsh environments (temperature, UV exposure, wind, rain), the presence of substantial cattle mass (affecting wireless communications), battery lifetime, and destructive animal behavior (potentially damaging instruments). This paper describes a framework for economically collecting high volume, high quality data on a wide range of conditions of individual head that can be leveraged to draw inferences about the dynamic behavior of the herd. A ruggedized Bluetooth cattle halter platform has been implemented the Radio Signal Strength Indicator (RSSI) of which was used for triangulation. Implementation of a proprietary antenna was required in order to provide high fidelity distance measurements as required for precise triangulation. The antennas were durable, held in position reliably on the halter and provided an isotropic radiation pattern as required to measure distances accurately from all directions. A correlation coefficient of r2 = 0.82 has been measured between signal strength and distance in all directions as required for high resolution cattle location. The Bluetooth SPI interface can be leveraged to wirelessly relay a wide variety of additional sensor measurements to a central computer from a number of feedlots the number of which can easily scale. Access to unprecedented quantities and qualities of herd data can be leveraged to dramatically improve operational efficiencies and realize precision agriculture.

Language

en

Provenance

Received from ProQuest

File Size

45 pages

File Format

application/pdf

Rights Holder

Luis Carlos Bañuelos Chacon

Download

Share

COinS