In Part 2 of this engaging episode of the Physical Activity Researcher Podcast, Dr. Olli Tikkanen continues the conversation with Professor Suresh Neethirajan about using Fibion SENS accelerometers and AI models to transform animal health monitoring. In this segment, Prof. Neethirajan delves into the data analysis techniques used to assess and improve animal welfare on farms. Watch or listen to Part 2 of the podcast here:
About Professor Suresh Neethirajan
Prof. Suresh Neethirajan, an expert in digital agriculture and animal health, is driving innovative, data-centered research at Dalhousie University. His work uses Fibion SENS to track animal movement in precise detail, providing valuable insights into health and welfare patterns. This real-time data supports the early detection of potential health concerns, setting new standards in proactive animal care.
How Fibion SENS and AI Work Together for Animal Welfare
Prof. Neethirajan explains how combining Fibion SENS data with other variables—such as milk production, milk quality, and facial images—creates a robust model for identifying health issues like lameness in dairy cows. By correlating accelerometer data with time-stamped health indicators, his team can accurately monitor movement patterns, even discerning nuances like walking speed and standing-to-lying ratios.
Building High-Accuracy Predictive Models
DenseNet and LSTM Models for Precision Analysis
In this part, Prof. Neethirajan reveals that his team has developed a DenseNet model paired with LSTM (Long Short-Term Memory) for highly accurate predictions of health outcomes. By using Fibion’s high-frequency data, they achieve prediction accuracy rates as high as 95%, allowing early intervention and improved animal care. This model, validated in studies on turkeys and chickens, also offers potential applications in other livestock.
High-Resolution Data for Improved Insights
Prof. Neethirajan emphasizes the value of high-resolution data collected by Fibion SENS for capturing “iceberg indicators” of animal pain and stress. The data’s frequency is adjusted for each species—for example, data collected every 15 seconds for chickens or every minute for dairy cows. This flexible approach ensures high-quality data without overwhelming storage, making Fibion SENS an ideal tool for diverse animal applications.
To explore these powerful applications of Fibion SENS and advanced AI models in animal health, watch or listen to Part 2 above.
Call to Action
📅 If you want to learn more about Fibion SENS, do not hesitate to book a video call with our expert Dr. Miriam Cabrita.
🔍 Order Fibion SENS Motion 3 Devices Test Package to get first-hand experience of the system.
