A S.M.A.R.T. approach to fly and tick surveillance in beef cattle production systems

The U.S. cattle industry loses an estimated $5.7 billion USD to fly and tick feeding annually. These ectoparasites blood feed on their hosts, causing blood loss, reduced feed efficiency, reduced milk quality and quantities, hide damage, and pathogen transmission. Diseases associated with these pests include pink eye, mastitis, theileriosis, and anaplasmosis. Traditional methods of fly and tick surveillance are time consuming and often unsafe for producers; therefore, we are aiming to automate that process. We hypothesize that the development of a neural network will provide stakeholders with a rapid, accurate, and safer method of detecting flies and ticks on cattle and will help to improve the overall efficiency and accuracy of pest detection when used in combination with manual inspection methods. The objective of this project is to develop a S.M.A.R.T. surveillance platform for fly and tick detection in cattle production systems using automated computer vision. A neural network is being developed using human-annotated images of flies and ticks on various host animals. The intended outcome of this project is to use the developed neural network to create a digital pest monitoring system that would allow users to upload images of their pest-infested cattle and obtain an instantaneous pest population estimate. By automating record keeping of pest populations, we aim to encourage producers to adopt a preventative approach to managing arthropod pests, laying the foundation for the development of IPM guidelines focused on reducing reliance on chemical control methods, improving animal welfare, minimizing environmental impacts, enhancing biosecurity, and improving productivity.