Towards understanding animal welfare by observing collective flock behaviors via AI-powered Analytics

Abstract

Animal farming has undergone significant transformation and evolved from small-scale businesses to large-scale commercial ventures. While maximizing productivity and profitability has always been a major concern in animal farming, during recent years there has been an increasing rise of concern regarding the welfare of the animals. In this context, the integration of artificial intelligence (AI) technologies offers immense potential for monitoring the well-being of chickens on farms and optimizing revenue streams simultaneously. Several works have integrated AI methodologies into everyday animal farming activities. Still, very few (if any) have proposed efficient and practical solutions that may facilitate farm owners in making impactful decisions regarding their business profitability and the welfare of the animals. In this direction, we propose a non-invasive chicken farm monitoring system that relies on onfield sound and video recordings integrated with sensory data acquired from the farm. The system consists of hardware that handles data acquisition and storage, a sensory data collection system and audio/video processing AI models. The last component of the system will be an inference engine that analyzes the collected data and infers useful facts about the flock's welfare and even psychological state.