Development of an IoT Platform for Stress-Free Monitoring of Cattle Productivity in Precision Animal Husbandry

Abstract

Smart animal husbandries require the adoption of dedicated tools to assess the contribution of each animal to the production process. The IoT platform presented in this article is a real-team monitoring system for voluntary weighing of cattle. To this end, the ISO 18000-6 standard is used for animal identification through an ultra-high-frequency radio link between a reader antenna and suitable ear tags. A customized data processing algorithm has been developed and embedded in the considered system. To demonstrate the effectiveness of the solution, extensive measurements have been carried out in a real-life environment. The proposed IoT platform is useful to farmers as a control tool for selection and breeding work.

Systematic Review on Internet of Things in Smart Livestock Management Systems

“…This paper lacks implementation details. Mirmanov et al [49] used Internet of Things-based techniques to monitor the weight of cows on farms. Ultra-high-frequency radio-frequency identification (UHF RFID) tags were used to identify cows.…”

A Review on Information Technologies Applicable to Precision Dairy Farming: Focus on Behavior, Health Monitoring, and the Precise Feeding of Dairy Cows

“…In dairy farming management, the use of RFID non-contact automatic identification technology can provide accurate individual data for each cow, which is of great significance for the orderly management and monitoring of the entire process of breeding links. Mirmanov et al (2021) developed automatic cattle weighing systems with RFID systems, and they have passed experimental tests and allow for assessing not only the dynamics in weight changes but also accurately displaying the weight of the animal [22]. However, the RFID device usually needs to be attached to an animal, which may be lost, removed, or damaged [23].…”

“…Deep-learning approaches with powerful feature extraction and image representation abilities also have been applied for cattle identification purposes [11]. These comprise a non-contact method of cow identification and exhibit higher recognition accuracy; they are represented by convolutional neural networks and can not only learn and classify the target in the image but also accurately predict the location of the target [22]. For instance, Kumar et al (2018) proposed a CNN-based approach to identify individual cattle using primary muzzle point images, and 98.99% accuracy was achieved [24].…”