AI-Based Automated Cleaning System Design for the Peruvian Poultry Industry

The poultry industry faces biosecurity challenges, particularly in cleaning and disinfecting crates used to transport live poultry, as these accumulate residues that foster pathogen spread. This study presents the design and simulation-based validation of a tunnel-type machine for crate cleaning, integrating high-pressure spraying, a YOLOv11-based computer vision module for real-time dirt detection, and a steam disinfection stage modeled through thermal simulations. The system optimizes water use, reduces washing time, and complies with SENASA biosecurity standards. In simulations, YOLOv11 achieved an F1Score of 95.71% using Stochastic Gradient Descent, providing a methodological basis for future physical implementation in industrial environments.