Skip to main content
ARS Home » Plains Area » Las Cruces, New Mexico » Range Management Research » Research » Publications at this Location » Publication #429312
Research Project: Knowledge Systems and Tools to Increase the Resilience and Sustainability of Western Rangeland Agriculture

Location: Range Management Research

Title: Architecture and implementation of a Precision Ranching Platform for Rangeland and Livestock Management

Author
item UTSUMI, SANTIAGO - New Mexico State University
item BAKIR, MEHMET - New Mexico State University
item PEREA, ANDRES - New Mexico State University
item Estell, Richard
item Cibils, Andres
item CAO, HUIPING - New Mexico State University
item Spiegal, Sheri
item McCord, Sarah
item Bestelmeyer, Brandon

Submitted to: Journal of Animal Science
Publication Type: Abstract Only
Publication Acceptance Date: 9/1/2025
Publication Date: 12/10/2025
Citation: Utsumi, S.A., Bakir, M., Perea, A.R., Estell, R.E., Cibils, A.F., Cao, H., Spiegal, S.A., McCord, S.E., Bestelmeyer, B.T. 2025. Architecture and implementation of a Precision Ranching Platform for Rangeland and Livestock Management. Journal of Animal Science. Supplement.

Interpretive Summary:

Technical Abstract: Applying adaptive grazing management on extensive rangelands could be time- and labor-demanding and challenging. However, advancements in virtual fencing, remote sensing, IoT sensor networks, and artificial intelligence offer new tools to streamline operations and make livestock and rangeland management more efficient. We developed a scalable IoT platform for Rangeland and Livestock Management using a LoRaWAN US915 protocol. The system is being deployed across approximately 1,000,000 acres of arid and semiarid rangelands spanning ten southwestern cow-calf operations in four states and monitoring approximately 1,000 head of cattle. Solar-powered gateways, positioned for optimal connectivity (15 km), collect millions of data packets from cattle tracking collars, water level sensors, rain gauges, and soil moisture probes, transmitting data in real-time to a network server. The architecture includes a Flask-based server with MongoDB for raw data storage, accessible via a web-based dashboard. Raw data is processed into actionable insights such as real-time visualizations of water levels and precipitation, while computer vision and machine learning models are applied online to accurately classify grazing, walking, and resting behavior (F1 = 0.94). Ensembles of neural networks efficiently detect anomalies (F1 > 0.85) related to calving, and eventual health issues or predation. The system integrates remote sensing tools from the Rangeland Analysis Platform (RAP), enabling enhanced vegetation and grazing monitoring at 16-day intervals. Efficient task execution is managed with Celery, Redis improves data caching and response times, and PostgreSQL stores processed data. A Django-based application dashboard allows users to manage accounts, access sensor data, and query RAP visualizations, while Docker ensures scalability. By integrating animal tracking, virtual fencing,and forage production, the platform facilitates applying adaptive grazing management decisions, optimizing resource use and rangeland resilience.