Autonomous mobile robot for disease detection and individual health monitoring of cattle in intensive dairy farms

Abstract

Intensive farming methods” used in modern day dairy farms have resulted in a need for close monitoring of cattle health and early identification of diseases. The labour shortage faced by dairy industry, globally, has left farmers with no option but to go for automation. Current semi-automated systems used for cattle health monitoring, lack a non-invasive, real-time information system for individual health monitoring of cattle. As an attempt to address these issues, this research was done with the objective of finding out the suitability and feasibility of an “Autonomous Mobile Robot” to monitor cattle individually and update the stakeholders with real-time information. The design of the system involved a robot, programmed to navigate from one cow to another while cattle are being fed in “feeding stations”. A visual-imaging camera and a thermographic camera were used as sensors to detect a set of physiological indicators corresponding to symptoms of popular cattle diseases. A “face recognition algorithm” was developed for cattle and images taken from both visual and thermographic cameras were processed to extract information about health condition of each cow. The information would then be communicated to the herdsmen via internet on a real-time, simplex communication system. Sensors used were capable of detecting two different physical parameters, namely, red color on facial area and body temperature of cattle. The algorithms were simulated using real-world imagery. Detection rates obtained through simulations proved the algorithms to be effective. Despite the high detection rates, some limitations which could hinder system performance such as poor barn conditions and inadequate system parameters were identified. However, as the barn conditions and design parameters could be modified and improved, the conclusion is that the mobile robot based system can fulfil the objectives of the research.