Machine Learning Assessment of Poultry Respiratory Health Using Housing Microclimate Data

Abstract

Introduction: Health concerns associated with the respiratory systems in poultry have been raised because of their implications in welfare, growth and productivity of the birds. The health and respiratory condition in the poultry are influenced by various environmental factors in poultry farms, such as temperature, humidity, air quality and ventilation. This paper has proposed an evaluation framework for assessing the respiratory health of poultry through the application of machine learning techniques in analyzing housing environment microclimate data. Environmental factors and health related factors were combined in the approach to capture the association with respiratory stress conditions. The microclimate data was preprocessed and analyzed through machine learning algorithms to classify and predict the health status of poultry. The findings indicated that machine learning algorithms efficiently capture the relationship between the variations in the microclimate and the responses of poultry respiratory system. The monitoring framework offers enhanced monitoring capabilities compared to observation-based monitoring approach. The research concluded that machine learning monitoring of housing environment could help to improve precision poultry farming.