Media Releases and Opportunities

It is an innovation sparked by a simple, yet quirky question: If cows could speak, would you listen and, if you did, could you decipher the meaning of their snorts and moos?

A researcher at Dalhousie University was curious and explored the possibility of understanding bovine communications in a bid to improve both animal welfare and farming. His research has resulted in a new app that can help farmers interpret their herd's sounds -- or 'emotional fingerprints' -- which could make the difference between a healthy herd and thousands of dollars in losses.

MooLogue bridges scientific research and farm-ready technology to interpret an animal's voice and act as an early warning system to help farmers save money, improve productivity and meet the rising demands of sustainable agriculture.

Dr. Suresh Neethirajan, who has a joint appointment in the faculties of Computer Science and Agriculture, and his research group created a comprehensive dataset of dairy cow vocalizations so they would have enough high-quality sound data to train AI models to detect changes in cow well-being before they became visible.

They used highly sensitive microphones and field recorders to capture more than 300 hours of cow vocalizations in milking parlours, feeding lanes, water stations and resting areas, logging not just the sound, but the time of day, activity, herd composition and even weather conditions.

The recordings ranged from the harmonic moos of a mother calling her calf and the sharp, urgent notes of cows that were uncomfortable to the low, conversational moos that seemed to bond the herd. They also revealed how socially and emotionally complex dairy cows are: they respond quickly to environmental changes and their vocal patterns shift accordingly.

The team trained deep learning models on the annotated audio, teaching them to distinguish between call types and link them to probable emotional states. That insight makes the app a valuable tool for farmers in identifying potential bovine health issues, lessening the chances of disease outbreaks or supply disruptions.

Dr. Neethirajan is available to discuss MooLogue and how it complements existing herd management systems by adding vocal patterns to traditional monitoring metrics.

MooLogue video: https://www.youtube.com/watch?v=ZuR7k09bb64
MooLogue app: https://apps.apple.com/us/app/moologue/id6749261442

Media contact:

Alison Auld
Senior Research Reporter
Dalhousie University
Cell: 1-902-220-0491
Email: alison.auld@dal.ca