Computational System to Support Bovine Nutritional Behavior

Introduction

The beef industry plays an important role in the economy worldwide with Brazil being responsible for 18% of the global exports in 2016 with a gross value of almost US$ 5 billion. In Brazil, beef production mainly occurs on tropical pastures, reducing production costs. However, the sector has several challenges related to efficiency and sustainability. In general, there is a growing demand for technologies that increase efficiency in animal management, nutrition and in the farm management (ABIEC, 2015; DA Silva, 2013; MAPA, 2016).

The need to understand the feeding process to identify cattle preferences requires the use of new technologies and adoption of best practices. This has led to the emergence of precision livestock farming, which can be defined as the management of livestock production using the principles and technology of process engineering, relying upon automatic monitoring of livestock and treating livestock production as a set of interlinked processes, which act together in a complex network. Therefore, precision livestock farming consists also of measuring different animal variables and modeling these captured data. The information resulting from these data can be used to monitor and control herds (Carvalho et al., 2009; Wathes et al., 2008).

In nutritional management, for example, there are several opportunities for using technology to generate information to improve decision-making. For grazing cattle, supplying nutrients is usually necessary in order to have economically viable production. These nutrients are provided in the form of a blend of many feed ingredients called supplement. The supplement has to be available for animals in troughs that cattle visit to feed daily (Fraga Filho; Resende, 2011) and the task of guaranteeing the availability and optimizing delivery can be greatly improved with precision livestock techniques.

The behavior pattern of cattle visits to the trough may generate information about the environment imposed on it. For example, sporadic visits may indicate an inadequate supplement, which causes the animals have less drive to consume or inadequate trough availability and prevents all animals in the group to consume the supplement (Fraga Filho; Resende, 2011). If the farm manager has an effective and quick tool to identify these inadequacies he can, in a shorter period of time, make decisions like modifying the supplement in favor of greater productivity.

In this work the researchers carried out a survey of the literature and in the national and international databases of patents and software registries to identify a computational system that could capture the visits of the animals to troughs located in pastures. Those systems should be able to send this data to an information system that could handle them properly and provide management reports. Computational solutions already commercially available for capturing and transmitting the data from the visits to the troughs have efficient results (Growsafe, 2004; Intergado, 2015), but they demand a high financial investment that extensive livestock farming cannot cope with. Authors could not find any specific information system that, based on the visit data, provided reports to evaluate the behavior of the animals about their visits to the supplementation trough.

This article aims to present a computational system composed of hardware and software elements for the precision livestock area to support the management of bovine supplementation. The hardware module collects data on the visits of the animals to the troughs on pasture and transmits these data to the software module, which processes them and generates reports to support decision-making.