Non-intrusive and effective positioning of precision livestock monitoring system for cattle

Abstract

Livestock farmers in Kenya face a number of challenges such as diseases and the inability to accurately identify oestrus and calving windows. These factors hinder productivity and lead to high livestock mortalities. Precision Livestock Farming systems are the solution to this, ensuring effective management of the livestock farming process. Existing systems have been able to monitor animal temperatures, location and movement of animals within the farms, and deliver this to the farmer visually, allowing them to observe the state of their animals in real-time and affordably. Positioning of sensors on livestock is critical in ensuring correct livestock data is collected. The placement of these sensors is dependent on three factors; the thermal windows on the animal’s surface, fastening of the sensor and the power supply. Real time data on sensor temperature readings from various parts of a cow’s body was obtained and analysed. The results were then compared alongside data from literature to come up with a preferred positioning of PLF sensor systems. By comparing the placement of the sensor on the cow’s leg, dewlap and harness, it was noted that the harness provided for a more suitable placement of the particular PLF sensor, allowing for continuous and accurate collection of data. Keywords—Precision livestock farming, machine learning algorithm, livestock, sensor positioning.