Performing Department
(N/A)
Non Technical Summary
US swine industry contributes 30% of domestic meat production, generating over $28 billion in gross cash receipts. Majority of swine are in confinements to enhance animal welfare and productivity. However, the existing sensor-based industry standard for pig confinement environmental control cannot guarantee optimal growth conditions and animal outcomes. The goal of this project, (P3), is to develop a pig-based and AI-powered proactive indoor environmental control and management system to optimze productivity, welfare, and sustainability. The project unites a trans-disciplinary team of subject matter experts in research, extension and education to accomplish five objectives (1) Use PLF techniques to collect feeding and drinking behaviors, activity budgets, and physiological measures under various growth conditions, and conduct manual data labeling for events classification, (2) Develop AI models to automatically detect pig outcomes and explore statistical-based interaction modeling of pig outcomes and environmental conditions, (3) Develop an augmented deep reinforcement learning-based AI system and integrate it with indoor environmental controls creating a proactive control and management system, (4) P3 proof-of-concept test to assess benefits in animal performance, welfare, energy and water usages, air emissions and waste production. (5) Integrate P3 principles into educational and extension resources and conduct training for swine production workforce and stakeholders. The long-term impact of this work is to enhance the sustainability of US animal production. It is expected that P3 will empower stakeholders with game-changing knowledge, tools, technology, and an interdisciplinary trained workforce.
Animal Health Component
0%
Research Effort Categories
Basic
50%
Applied
50%
Developmental
0%
Goals / Objectives
The overarching goal of the project is to develop a pig-based and AI-powered proactive indoor environmental control and management system (a.k.a. Proactive Pig Production, P3) to optimize productivity, welfare, and sustainability.We will pursue the following supporting integrated objectives to achieve the project goalUse PLF techniques to collect feeding and drinking behaviors, activity budgets, and physiological measures of pig outcomes (i.e., body surface T, respiration rate) under various growth conditions and conduct manual data labeling for events classification.Develop offline AI models to automatically detect pig outcomes and explore statistical-based interaction modeling of pig outcomes and environmental conditions.Develop an augmented DRL-based AI system using PLF monitoring data and integrate it with indoor environmental control to create a proactive control and management system, P3.Proof-of-concept test of P3 to assess its benefits in animal performance, welfare, energy and water usages, air emissions and waste production.Integrate P3 principles into the development of formal and informal educational and extension resources and experiences; and conduct transdisciplinary training for swine production workforce and stakeholders.
Project Methods
This project unites a trans-disciplinary team of subject matter experts in research, extension and education to accomplish five objectives (1) Use PLF techniques to collect feeding and drinking behaviors, activity budgets, and physiological measures under various growth conditions, and conduct manual data labeling for events classification, (2) Develop AI models to automatically detect pig outcomes and explore statistical-based interaction modeling of pig outcomes and environmental conditions, (3) Develop an augmented deep reinforcement learning-based AI system and integrate it with indoor environmental controls creating a proactive control and management system, (4) P3 proof-of-concept test to assess benefits in animal performance, welfare, energy and water usages, air emissions and waste production. (5) Integrate P3 principles into educational and extension resources and conduct training for swine production workforce and stakeholders.