Ai-driven smartphone app promises faster Salmonella detection in poultry farms

A new AI-based mobile application developed by the University of Georgia could revolutionise Salmonella surveillance in poultry production by enabling rapid, on-site diagnosis through smartphone images, potentially preventing outbreaks earlier.

Artificial intelligence could give poultry producers a faster way to flag Salmonella risk by analysing smartphone images of droppings, according to a University of Georgia study published in Computers and Electronics in Agriculture.

The system, called “Salmonella Detector: A Mobile Application for On-site Broiler Salmonella Infection Diagnosis”, uses deep learning and computer vision to sort faecal images into PCR-positive or PCR-negative categories. Led by Guoming Li of the university’s Department of Poultry Science and Institute for Artificial Intelligence, the project was designed as a low-cost screening tool that could help producers spot problems sooner and act before infection spreads through a flock.

At present, Salmonella surveillance usually depends on collecting samples and sending them to a laboratory, a process that can take several days. The researchers argued that a field-ready tool could improve decision-making by providing quicker, on-farm indications of risk.

The team tested thousands of images gathered from poultry farms in the United States and Africa. Performance was strongest when the model was trained and tested within the same region, but cross-regional testing still reached up to 85% accuracy, suggesting wider potential while also underscoring the effects of differing environments, management systems and image characteristics.

To improve practicality, the researchers combined image analysis with tree-based machine learning and built prototype web, iOS and Android applications. They said the software has been made publicly available to support further research and development.

The authors stressed that the tool is intended to complement laboratory diagnostics, not replace them, and said broader validation is still needed before routine use. The work was supported through the U.S. Poultry & Egg Association’s research programme.

"This news service is powered by an AI tool that curates and enhances the content related to animal health. The articles provided are for informational purposes only and should not be considered professional or veterinary advice. The content presented does not reflect the views or opinions of Stonehaven Analytics."