Poultry Farm Intelligence: An Integrated Multi-Sensor AI Platform for Enhanced Welfare and Productivity

TL;DR

PoultryFI addresses the lack of affordable, integrated tools for continuous welfare and productivity monitoring on small to medium poultry farms. It presents an end-to-end platform that combines six modules—Camera Placement Optimizer, Audio-Visual Monitoring, Analytics & Alerting, Real-Time Egg Counting, Production & Profitability Forecasting, and a Recommendation Module—validated on edge hardware. Key contributions include offline camera layout optimization using CMA-ES and MAP-Elites, unsupervised audio anomaly detection via Conv-DAE, edge-based egg counting with EfficientDet-Lite0 on Raspberry Pi 5, short-term environmental forecasting, and prescriptive actionable guidance. Field trials in commercial settings demonstrate 100% egg-count accuracy on Pi 5, robust anomaly detection, and reliable short-term forecasts, illustrating the platform’s practicality, scalability, and potential for real-world adoption. Overall, PoultryFI enables proactive flock welfare and profitability management through multi-modal sensing, edge AI, and end-to-end integration.

Abstract

Poultry farming faces increasing pressure to meet productivity targets while ensuring animal welfare and environmental compliance. Yet many small and medium-sized farms lack affordable, integrated tools for continuous monitoring and decision-making, relying instead on manual, reactive inspections. This paper presents Poultry Farm Intelligence (PoultryFI) - a modular, cost-effective platform that integrates six AI-powered modules: Camera Placement Optimizer, Audio-Visual Monitoring, Analytics & Alerting, Real-Time Egg Counting, Production & Profitability Forecasting, and a Recommendation Module. Camera layouts are first optimized offline using evolutionary algorithms for full poultry house coverage with minimal hardware. The Audio-Visual Monitoring module extracts welfare indicators from synchronized video, audio, and feeding data. Analytics & Alerting produces daily summaries and real-time notifications, while Real-Time Egg Counting uses an edge vision model to automate production tracking. Forecasting models predict egg yield and feed consumption up to 10 days in advance, and the Recommendation Module integrates forecasts with weather data to guide environmental and operational adjustments. This is among the first systems to combine low-cost sensing, edge analytics, and prescriptive AI to continuously monitor flocks, predict production, and optimize performance. Field trials demonstrate 100% egg-count accuracy on Raspberry Pi 5, robust anomaly detection, and reliable short-term forecasting. PoultryFI bridges the gap between isolated pilot tools and scalable, farm-wide intelligence, empowering producers to proactively safeguard welfare and profitability.

Figures (10)

• Figure 1: PoultryFI modules grouped by their primary impact on Farm Operations, Animal Welfare, or both. The Real-Time Egg Counting and Production and Profitability Forecasting modules are designed to streamline farm management and planning. The Audio-Visual Monitoring Module focuses specifically on monitoring animal welfare indicators. The Camera Placement Optimizer, Analytics and Alerting Module, and Recommendation Module contribute to both areas by transforming welfare-related data into actionable insights that support more efficient and responsive farm operations.
• Figure 2: Overview of the PoultryFI System. Grey rectangles indicate the six core modules; white, rounded rectangles represent data sources. Solid arrows show the flow of information, while the dotted arrow indicates an offline process. The Camera Placement Optimizer determines optimal camera positions to maximize poultry house coverage and is executed offline. The Real-Time Egg Counting module supports egg counting by analyzing video data when farmers log production. The Audio-Visual Monitoring Module (AVMM) processes synchronized video and audio streams to generate welfare indicators, including motion score, audio anomaly score, and feeder status. The Analytics and Alerting Module (AAM) combines data from AVMM and environmental sensors (e.g., temperature, humidity) to issue alerts about short-term deviations in flock activity or environmental conditions. The Production and Profitability Forecasting Module (PPFM) uses production data, environmental data, and AVMM indicators to generate longer-term forecasts on poultry house productivity and profitability. Finally, the Recommendation Module synthesizes alerts from AAM, forecasts from PPFM, and external weather data to provide actionable recommendations for farm management.
• Figure 3: The profile of Temperature on 20/04/2025 for multiple day intervals
• Figure 4: The profile of Humidity on 20/04/2025 for multiple day intervals
• Figure 5: (Top) The Initial Manual Solution. (Middle) Solution provided by CMA-ES that was used to deploy the cameras in the poultry house. (Bottom) Solution provided by MAP-Elites similar to the deployed solution. Figures adapted from Mosphilis2025.