Developing and validating a S.M.A.R.T. surveillance platform for fly and tick detection in beef cattle

The United States cattle industry loses an estimated $5.7 billion USD to fly and tick feeding annually. These ectoparasites blood feed on their bovine 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 bovine 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 and validation of a neural network will provide stakeholders with a rapid, accurate, and safe method of detecting flies and ticks on beef cattle and will help to improve the overall efficiency and accuracy of pest detection when used in combination with manual (e.g., visual and hands-on) inspection methods. The objective of this project is to develop and validate a S.M.A.R.T. (sensors, monitoring, analysis, and reporting technologies) surveillance platform for fly and tick detection in beef cattle production systems using an automated computer vision system. A neural network will be developed using ~1,000 volunteer-submitted and human-annotated images of flies and ticks on cattle. To evaluate the reliability of the network, we will use additional images and counting results of multiple human scorers. The intended outcome of this project is to use the developed neural network to create an effective, user-friendly digital pest monitoring system (app and/or web-based platform) that would allow users to upload images of their pest-infested cattle and obtain an instantaneous pest population estimate.