Machine Learning-Based Sensor Data Fusion for Animal Monitoring: Scoping Review

Abstract

The development of technology, such as the Internet of Things and artificial intelligence, has significantly advanced many fields of study. Animal research is no exception, as these technologies have enabled data collection through various sensing devices. Advanced computer systems equipped with artificial intelligence capabilities can process these data, allowing researchers to identify significant behaviors related to the detection of illnesses, discerning the emotional state of the animals, and even recognizing individual animal identities. This review includes articles in the English language published between 2011 and 2022. A total of 263 articles were retrieved, and after applying inclusion criteria, only 23 were deemed eligible for analysis. Sensor fusion algorithms were categorized into three levels: Raw or low (26%), Feature or medium (39%), and Decision or high (34%). Most articles focused on posture and activity detection, and the target species were primarily cows (32%) and horses (12%) in the three levels of fusion. The accelerometer was present at all levels. The findings indicate that the study of sensor fusion applied to animals is still in its early stages and has yet to be fully explored. There is an opportunity to research the use of sensor fusion for combining movement data with biometric sensors to develop animal welfare applications. Overall, the integration of sensor fusion and machine learning algorithms can provide a more in-depth understanding of animal behavior and contribute to better animal welfare, production efficiency, and conservation efforts.

Automated monitoring of brush use in dairy cattle

“…We tested and compared three commonly used supervised ML models [ 16 , 32 , 33 ]: 1) Random Forest (RF) [ 34 ] operates through a collection of decision trees, each providing its input on the data. By aggregating the decisions of multiple trees, it achieves a more accurate and stable prediction.…”

“…For this task we used Scikit-learn’s multi target classification which consists of fitting one classifier per target, a strategy for extending classifiers that do not support multi-target classification [ 31 ]. We evaluated 3 commonly used ML algorithms [ 16 , 32 , 33 ]: 1) Multi-Layer Perceptron (MLP) [ 37 ] is a kind of neural network that consists of multiple layers through which data is processed, allowing the model to learn complex patterns. It is adept at tasks where the relationship between input data and the output is intricate, making it suitable for complex classification problems.…”

Comparison of Machine Learning Tree-Based Algorithms to Predict Future Paratuberculosis ELISA Results Using Repeat Milk Tests

“…While ML-guided decision support systems have been used in human medicine since the 1970s, their application in veterinary medicine has been few and far between [10]. Most applications of AI in the veterinary and animal husbandry sector have been focused on solving animal health and welfare problems through analyzing animal behavior [11,12] with very few studies addressing disease detection [11]. In a review by Basran and Appleby [13], the authors highlight the unmet potential for AI in veterinary medicine; however, the published work on AI in veterinary medicine has mainly been centered on medical image analysis (e.g., radiographs, CT, MRI, ultrasound images, and teat images on dairy cattle).…”

Analyzing Cattle Activity Patterns with Ear Tag Accelerometer Data

“…Commercial ear tag accelerometers have also been used to identify behaviours and activity levels of grazing cattle [ 22 , 23 ]. The success of the associated data-driven solutions depends largely on the quantity and quality of the available ground-truth data used for model training as well as the features extracted from the accelerometer data [ 24 , 25 ].…”