Explainable AI for Livestock Disease Detection: An Integrated ML/DL Framework
B Ajay Vardhan *
Sreenidhi Institute of Science and Technology, Yamnampet, Ghatkesar, Hyderabad, Telangana-501301, India.
K.P Rushil Phanindra
Sreenidhi Institute of Science and Technology, Yamnampet, Ghatkesar, Hyderabad, Telangana-501301, India.
K Sumith
Sreenidhi Institute of Science and Technology, Yamnampet, Ghatkesar, Hyderabad, Telangana-501301, India.
T. Sirisha
Sreenidhi Institute of Science and Technology, Yamnampet, Ghatkesar, Hyderabad, Telangana-501301, India.
V. Kakulapati
Sreenidhi Institute of Science and Technology, Yamnampet, Ghatkesar, Hyderabad, Telangana-501301, India.
*Author to whom correspondence should be addressed.
Abstract
Livestock diseases lead to significant economic loss and threaten food security. With the increasing demand for dairy and meat products, maintaining animal health has become a critical global priority. Although farmers and agricultural workers often lack deep technical understanding of data processing, modern AI and ML technologies are now central to early disease detection in livestock. Interpretable Machine Learning (IML) and Explainable AI (XAI) provide opportunities to build trust by making model predictions transparent and understandable. This article explores XAI and IML approaches for health monitoring in farm animals, offering insights into early symptom recognition through sensor data and image analysis. XAI integrates CNN-based visual diagnostics and real-time sensor stream interpretation, while IML utilizes SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations) for symptom pattern explanation and decision support. Experimental results using publicly available datasets of livestock behavior and visual symptom records demonstrate that XAI/IML-based systems provide farmers and veterinarians with clear, actionable insights to enhance livestock welfare and productivity.
Keywords: Explainable AI, IML, livestock, disease detection, SHAP, LIME, CNN, animal health, farmers, prediction, sensor data