As of 2013, there was an estimated 24 billion livestock globally (http://faostat.fao.org). The nature of the large scale livestock raising practices and increasing pressures of raising these animals in a humane, effective and safe way pose challenges to the owners, consumers and society at large (Fitzgerald, 2015a; 2015b). Meanwhile, the costs of wearable technology and cloud computing are decreasing, the predictive machine learning models are improving, thereby opening the potential for addressing these gaps via means of technology. This thesis provides a detailed overview on societal, commercial and technology issues regarding the livestock monitoring business and presents ways in which technology can address them. We discuss topics ranging from the economic, ethical, regulatory an...[ Read more ]

As of 2013, there was an estimated 24 billion livestock globally (http://faostat.fao.org). The nature of the large scale livestock raising practices and increasing pressures of raising these animals in a humane, effective and safe way pose challenges to the owners, consumers and society at large (Fitzgerald, 2015a; 2015b). Meanwhile, the costs of wearable technology and cloud computing are decreasing, the predictive machine learning models are improving, thereby opening the potential for addressing these gaps via means of technology. This thesis provides a detailed overview on societal, commercial and technology issues regarding the livestock monitoring business and presents ways in which technology can address them. We discuss topics ranging from the economic, ethical, regulatory and healthcare aspects surrounding the commercial viability of animal monitoring to hardware systems and algorithms that can support the livestock monitoring functions in theory as well as through case studies applied specifically on horse monitoring – in practice. This thesis therefore documents a mix of academic and practical knowledge of what it takes to build a functional and commercially viable system for animal most notably equine monitoring purposes using machine learning approaches.