Monitoring and Characterizing Micro- and Macro-Environment for sows and piglets in Experimental Swine Farrowing Building

Modern swine production is rapidly evolving with the integration of precision technologies to meet the growing food demand. As this evolution progresses, researchers have focused on addressing concerns related to animal welfare, health, and the challenges of production environments. However, understanding swine micro- and macroenvironments remains a significant challenge due to the complex interactions between animals, manure, ventilation systems, and pollutant gas emissions such as ammonia (NH₃), methane (CH₄), hydrogen sulfide (H₂S), and carbon dioxide (CO₂). Despite decades of research, there is a lack of real-time, high-resolution environmental monitoring systems that are both scalable and cost-effective for long-term use in swine facilities. Furthermore, limited studies have characterized microenvironment variations which may correlate to varying pig performance across pen. This dissertation addressed key gaps by using innovative methods to study micro- and macroenvironmental conditions in swine farrowing facilities.

Over four project trials across two years, the system demonstrated robust performance under harsh barn conditions. It achieved 99% data completeness from 100+ sensors tracking 20+ variables such as gas concentrations, ventilation, temperature, humidity, and pig activity, at 1 Hz frequency. Pig activity data, collected using an innovative PID sensor method, showed strong agreement with visual observations and video analysis, confirming its high accuracy and reliability. A novel low-cost airflow sensor correlated strongly with room ventilation (R² > 0.98), providing an innovative solution for real-time fan monitoring. However, challenges remain in sensor maintenance and long-term durability.

Analysis of high-frequency data revealed clear diurnal patterns in pig activity, temperature, humidity, and gas concentrations. At the room-level, thermal fluctuations were driven by animal activity, heater operation, and ventilation. Pollutant gas emissions increased by 7.8%–29% during sow move-in and decreased by 0.38%–14.2% during move-out. At the pen-level, significant variations were observed across the 12 pens, even during the empty barn period, with NH₃, CH₄, and H₂S concentrations varying by up to 10.1 ppm, 22.6 ppm, and 77.1 ppb, respectively. Relative humidity and temperature also fluctuated across pens by up to 3.41% and 0.82°C. These variations were further amplified by manure agitation events such as water leakage, power washing, and manure draining, resulting in H₂S spikes of up to 300%. Manure agitation, animal presence, and outdoor weather were identified as key drivers of fluctuations in gas levels, RH, and temperature.

This study presents scalable technology for real-world application and provides a deeper understanding of swine environmental dynamics. It emphasizes the importance of high-frequency, real-time monitoring to support data-driven decision-making in animal environment management, gas level control, and sustainable production. Future advancements should focus on expanding the system’s application to larger scales and incorporating AI-driven automation to further validate its feasibility and optimize swine production environments.