• Title: “Development of methods to characterize cattle behavior and social interactions” ; INRAE.
  • PhD student: Joseph Alyndrée
  • Affiliated unit: INRAE, Bioepar
  • Co-supervision:
    • Aurélien Madouasse (INRAE)
    • Antoine Cornuéjols (INRAE)
    • Christine Martin (AgroParisTech)
  • Doctoral school: Doctoral School of Plant, Animal, Food, Sea, and Environment
  • Project duration: 2023 - 2026

Project summary:

This thesis explores the use of sensors to monitor dairy cow behavior and machine learning techniques to better manage animal health, particularly lameness (a common and painful condition).
Thesis objectives:

• Behavioral Characterization: Use 3D accelerometers and machine learning methods to detect lameness by studying the rising and lying down movements of cows.
• Social Interactions: Study the consequences of lameness on interactions among cattle by combining different sensors and machine learning techniques.

Methodologies:

• Data Collection: Record cattle behaviors on dairy farms by equipping them with collars containing sensors.
• Machine Learning: Use this data to predict behaviors by segmenting time series and applying algorithms or deep learning methods to identify specific patterns.
• Association Analysis: Evaluate the links between behaviors, social interactions, and lameness.