Lead researcher James Chen is developing an acoustic data-driven tool to help enhance animal welfare and lower methane emissions in precision livestock farming. Photo courtesy of James Chen.
Virginia Tech researchers are deciphering cow vocalizations using acoustic data and machine learning, with the goal of improving animal welfare and lowering methane emissions via precision livestock farming. James Chen, an animal data sciences researcher and assistant professor, is utilizing a $650,000 grant from the USDA’s National Institute of Food and Agriculture to create an acoustic, data-driven tool for analyzing cow vocalizations for indicators of stress or sickness.
Cows convey their feelings via vocalization, and we must pay attention to what they are saying. Sound data may be obtained from cows individually and constantly, making it superior than video or other techniques of observing cows’ emotions and health. By combining auditory data with biological and visual clues, we may be more objective in our analysis of behavior.
Chen and his co-investigator, Virginia Cooperative Extension dairy scientist and associate professor Gonzalo Ferreira, want to gather audio data from cows, calves, and beef cattle on the grassland. They will next use machine learning to evaluate and categorize hundreds of audio data points, as well as interpret cow vocalizations like mooing, munching, and burping for indicators of stress or disease.
Chen and Ferreira are especially interested in discovering the vocal patterns used by cows to indicate discomfort. By monitoring the frequency, amplitude, and length of cow moos and vocalizations and matching the sound data with saliva cortisol samples, scientists may determine if cows are under no stress, moderate stress, or severe stress, and begin to interpret their “language.”
The collected data will be made accessible via an open-source, web-based tool for scientists, producers, and the general public. The ultimate objective is to use this methodology on a broader scale, creating a public dataset to guide legislation and laws.
