Using broiler sound frequency to model weight

Chicken weight provides information about growth and feed conversion in order to identify deviations from the expected homogeneous growth trend of the birds. Precision Livestock Farming (PLF) can support the farmer through the use of sensors, cameras and microphones. Previous studies showed a significant correlation (p<0.001) between the frequency of vocalisation and the age and weight of the broiler. In this study, recordings were made in an automated, non-invasive way through the entire life of the birds, to evaluate the frequency variation of the sounds emitted during production cycles. In total, sound data collected during 8 production cycles (in an intensive broiler farm – 30,000 birds reared per round) were analysed. Sound data were manually and automatically compared with the weight of the birds automatically measured. Sound analysis was performed based on the amplitude and frequency of the sound signal in audio files recorded at farm level. The aim of this study was to sample automatically broiler vocalisations under normal farm conditions, to identify and model the relation between animal sounds and growth trend, and develop a tool to automatically detect the growth level of the animals based on the frequency of the vocalisation. The model used to predict the weight as a function of the Peak Frequency (PF) confirmed that the animal weight could be predicted by the frequency analysis of the sounds emitted at farm level although a more accurate editing of the audio file is necessary.