Complex event processing for effective poultry management in farmlands with predictive analytics

Abstract

Occurrence of an event may influence one or more events. Complex event processing (CEP) is a field focused on capturing real time events, analysing inter-dependences among myriad of events and arrive at insights. This study is made on the real time events in a poultry farm by applying CEP driven predictive analytics over both historical data and current data to forecast activities, behaviours and trends. A system was designed wherein individual birds are RFID tagged, monitored and sensor data for moisture content, light, time, weather was collected in cloud and analysed. Proposed model applies K-means clustering for behaviour patterns analysis. Machine learning algorithms were used to capture varied complex interactions influencing poultry well-being. Study was made on a farm with 846 country chickens, wherein prediction algorithms have enabled prediction of unusual behaviour patterns as well as foresee disease outbreak amongst chicken in the farm with an accuracy of about 78%.