Ivan Andonovic
Data-driven Machine Learning Precision Livestock Farming Technologies and Applications
Biography
Professor of Broadband Networks, graduated with a BSc and PhD in Electronic and Electrical Engineering from the University of Strathclyde. He has edited two books, authored/co-authored six book chapters, over 380 journal and conference papers and secured funding in excess of £15M. He was Topical Editor for the ‘IEEE Transactions on Communications’, Technical Programme Co-Chair for the ‘IEEE International Conference in Communications (ICC07)’, co-founder, Director and Chief Technology Officer of Kamelian Ltd., a technology start-up focusing on the manufacture of advanced semi-conductor devices and Silent Herdsman Ltd a company providing a range of cloud-based precision livestock health services.
Professor Andonovic’s research interests center on the development of broadband networks, distributed wireless sensor systems, Internet-of-Things and data-driven applications/services.
Contact Details:
Email: i.andonovic@strath.ac.uk
Mobile No: +447714981798
Office No: +44141 548 2537
https://www.strath.ac.uk/staff/andonovicivanprof/
Abstract
The Food and Agriculture Organisation (FAO) of the United Nations predicts that the global population will grow to around 10billion by 2050 and consequently food production must increase by up to 70% to meet that need. The target has to be achieved in spite of the limited availability of arable lands, the increasing need for fresh water (agriculture consumes 70% of the world’s fresh water supply) and other less predictable factors, such as the impact of climate change which leads to variations to seasonal events in the life cycle of plant and animals. Furthermore, agriculture faces a range of additional challenges from new pests and diseases that compromise output quality, generate harmful residues (drugs within the food chain) which in turn necessitate increased pollution management. In parallel, the growing trend of increasing farm sizes, and taking into account the remoteness of the farming community, translates into a migration from established practices based on visual inspection to more automatic techniques capturing data from plants/animals dispersed across a wide spectrum of locations.
Precision Livestock Farming (PLF) is thus core to satisfying the ever increasing world-wide demand for good quality products whilst heavily reducing environmental load and resource use. The pressing need to secure food supplies ensures that the adoption of technology-enabled solutions and applications/services will continue to gather pace. Increasingly on-farm dairy systems (heat detection, milk analysis, feed management etc) are being deployed through Cloud-based implementations, releasing the potential to provision a range of applications/services that bring benefit throughout the entire supply chain. The integration of multiple data streams yields significantly more value because a single indicator may return to normal over a period of time. The integrated data can be analysed to determine correlations between input/output parameters for each individual animal forming the basis for a range of services informed by the relationship between both and dissemination of alerts through multiple channels.
The paper details the features of a platform that implements PLF management strategies for the dairy industry. The platform elements comprise robust, high node count sensor networks gathering data from individual animals and a cloud based software environment that manages on-farm data and pro-actively alerts the farmer, real time, of key operational and management interventions. The principle is that if the needs of animals at the individual level are properly defined and met, then the needs of farmers and downstream stakeholders including consumers follow. The more precisely that needs are met, the less waste there is in the system, resulting in greater economic and environmental benefits. In turn, the creation of new business models based on provisioning a range of services to livestock farmers becomes possible, promoting the easy uptake of technology to all in the supply chain.
The platform is scalable in terms of handling multiple data streams, able to manage farms increasing in size including hybrid environments and support remote farms harnessing the growth of the connected world. The platform captures data from cows located in different areas, collates this information and presents it to a mix including the farmers/herdsmen, veterinarians, feed specialists over various mediums, such as smartphone, home computer or within the parlour.
Keywords: Data-Driven Applications/Services; Distributed Intelligent Sensor Networks and Applications; Internet of Things Technologies; Low Power Wireless Connectivity; Edge Computing; Machine Learning; Accelerometers; Precision Farming
