AI-Supported Detection of Footpad Lesions in Broiler Chickens

Initial Situation and Objectives

Footpad lesions are among the key animal welfare issues in broiler chicken production. They manifest as lesions and inflammation on the footpad and can lead to pain, restricted mobility, and—in severe cases—chronic impairments. The condition of the footpads is considered a well-established animal welfare indicator because it provides insights into housing climate, litter quality, and overall management practices.

Currently, assessment is largely carried out manually. This process is time-consuming, labor-intensive, and requires specialized expertise. In addition, it is subject to subjective influences and does not allow for continuous monitoring. The aim of the project is to develop an automated, image-based system for the objective detection and evaluation of footpad lesions in broiler chickens as a basis for improved animal welfare monitoring.

Technological Approach

The project develops AI-supported image analysis methods for detecting and classifying footpad lesions. The system is based on image data collected under real poultry house conditions.

Data acquisition is carried out using specially developed hardware designed for controlled individual animal handling and capturing the underside view of the footpads. Based on this data, the system will:

• automatically detect changes and lesions on the footpad
• classify the severity of the lesions
• derive objective and reproducible assessment metrics

The approach is designed for robust application under practical farming conditions and aims to reduce the need for manual assessment.

Benefits, Application, and Exploitation

The automated analysis replaces or complements manual animal welfare assessments through visual, data-driven monitoring. This allows abnormalities to be identified at an early stage and management measures to be implemented more effectively.

In the long term, the system enables the development of comprehensive, automated animal welfare monitoring in poultry farming. The objective data supports operational decision-making, contributes to improved flock health, and can reduce economic losses caused by delayed interventions.

Practical Environment

Development and testing are carried out in collaboration with Novu.track GmbH, the hardware developer of the data acquisition solution, as well as in additional real-world poultry housing environments.