Researchers associated with Semear Digital, Embrapa, the University of São Paulo and the Federal University of Grande Dourados have tested drone-based monitoring of feedlot cattle in Mato Grosso do Sul, Brazil. The team followed a group over 112 days, flying at roughly 15 metres to capture images.

Computer-vision models detected animals, segmented their bodies and extracted measurements such as length and width. These observations were used to model growth over time and identify changes in growth rate that could inform management decisions.

The study concerns advanced development rather than a ready commercial weighing service. Researchers identify further validation, adaptation beyond Nelore cattle and an industrial partner as next steps. A growth-curve turning point alone does not establish the economically optimal sale date for every farm.