CURRENT PERCEPTIONS, USE, AND FIELD IMPLEMENTATION OF PRECISION POULTRY FARMING TECHNOLOGIES

The agricultural sector is experiencing a profound shift driven by technological advancements and the imperative for sustainable food production. Within this context, Precision Livestock Farming (PLF) has emerged as a transformative approach to animal management (Berckmans, 2017), leveraging data-driven decision-making, automation, and monitoring systems to enhance livestock production efficiency while minimizing resource usage and environmental impact. This thesis delves into a comprehensive examination of PLF as applied to broiler farming, particularly focusing on current usage perceptions, needs, and its potential to revolutionize the poultry industry.

Research in Precision Livestock Farming has steadily increased since 2000, with a growing percentage of publications focusing on poultry. However, a significant portion of this literature centers on prototype products, with limited exploration of PLF perceptions. The primary objective of this thesis is to gauge the current perceptions, usage, and needs of PLF systems among academia. An online survey was conducted, targeting Animal Scientists, Engineers, and Veterinarians. Results indicated a consensus among poultry researchers regarding the future potential of PLF to enhance poultry production. While most respondents were familiar with PLF terminologies (92%), only about half reported current utilization of PLF technologies (55%). Cost and usability emerged as the primary barriers to implementation, cited by 93% and 88% of respondents, respectively. Notably, opinions varied based on experience with PLF technologies, with current users expressing greater confidence in PLF's potential to expand research opportunities.

The thesis also includes a field implementation project, testing the Scout® PLF system in a commercial broiler house. Scout® is a rail-mounted robot system designed for mortality detection and environmental monitoring. Regression analysis revealed strong linear relationships between Scout® measurements and reference sensors for temperature (R² = 0.88) and relative humidity (R² = 0.86), albeit with discrepancies outside expected sensor accuracy. Mortality detection accuracy was modest (15%), with unexpected clustering between feed and water lines. Bird occlusion was identified as a contributing factor, supported by experiments showing higher detection rates for birds within pens. In summary, addressing barriers such as cost, usability, and improving environmental monitoring systems are critical for the widespread adoption and success of PLF in the industry.