Quantification of grass-severing bites performed by grazing cattle using halter-mounted accelerometers and machine learning

Aug 5, 2026

Upload by Eden

Grasslands underpin sustainable food systems, and improving understanding of cattle grazing behaviour supports better pasture management and biodiversity outcomes. This study builds a two-phase machine learning methodology to detect grass-severing bite events from neck-mounted inertial measurement unit signals. Behaviour is first classified into mutually exclusive “ingestion” versus “other”, then bite counts are regressed during ingestion periods. Experiments used nine dry cattle with video-tagged training data to optimize window segmentation and data splits, achieving high ingestion classification accuracy and low bite-count error in low-switch sessions.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages15

Reading Time38 min

AccessFree to download

Introduction

Sustainable livestock farming and grasslands

Grass-severing bites at the plant–animal interface

See discussions, stats, and author profiles for this publication at: [https://www. researchgate. net/publication/382750542](https://www. researchgate. net/publication/382750542) Quantiﬁcation of grass-severing bites performed by grazing cattle using halter-mounted accelerometers and machine learning Article in Smart Agricultural Technology · July 2024 DOI: 10.1016/j.atech.2024.100522 CITATIONS 0 READS 69 5 authors, including: Nicolas Tilkens University of Liège 2 PUBLICATIONS 0 CITATIONS Frederic Lebeau University of Liège 135 PUBLICATIONS 2,570 CITATIONS Ali Siah Higher Institute of Agriculture-Junia 115 PUBLICATIONS 1,059 CITATIONS Andriamasinoro Lalaina Herinaina Andriamandroso Junia 16 PUBLICATIONS 492 CITATIONS All content following this page was uploaded by Nicolas Tilkens on 23 September 2024. The user has requested enhancement of the downloaded file. Smart Agricultural Technology 8 (2024) 100522 Contents lists available at ScienceDirect Smart Agricultural Technology journal [homepage:](homepage: www.journals.elsevier.com/smart-agricultural-technology)[ www.journals.elsevier.com/smart-agricultural-technology](homepage: www.journals.elsevier.com/smart-agricultural-technology) | Quantification of grass-severing bites performed by grazing cattle using halter-mounted accelerometers and machine learning | | | | | --- | --- | --- | --- | | N. Tilkensa, b, * , J. Bindelleb, c , F. Lebeau d , A. Siaha , A.L.H. Andriamandrosoa
a Junia, Universit´e de Lille, Joint Research Unit 1158 BioEcoAgro, 2 Rue Norbert S´egard, BP 41290, Lille F-59014, France
b AgricultureIsLife, TERRA Teaching and Research Center, Gembloux Agro-Bio Tech, University of Li`ege, Passage des D´eport´es 2, Gembloux 5030, Belgium c Recision Livestock and Nutrition Unit, AgroBioChem, Gembloux Agro-Bio Tech, University of Li`ege, Passage des D´eport´es 2, Gembloux 5030, Belgium d Precision Agriculture Unit, TERRA Teaching and Research Center, Gembloux Agro-Bio Tech, University of Li`ege, Passage des D´eport´es 2, Gembloux 5030, Belgium | | | | | A R T I C L E I N F O | | A B S T R A C T | | | Keywords: Cattle Livestock Behaviour
Inertial measurement unit Bite quantification Machine Learning | | Grasslands represent a key element of agroecosystems for sustainable food systems. A better understanding of the grazing behaviour of domestic herbivores is essential to support innovations for grassland management and define grazing practices that support rather than enter into conflict with biodiversity. A key component of the grazing process is the grass-severing bite by which the herbivore collects forage from a pasture. How often, where, and when such bites are performed are relevant indicators of the grazing behaviour of cattle and could be used as indicators to guide farmers in pasture management. In this work, we developed a methodology to create a Machine Learning (ML) model for identifying grass-severing bite events from the Inertial Measurement Unit (IMU) signals of a sensor placed on the neck of cows. The two-phase process consisted of classifying every period of behaviour of cattle into two mutually exclusive behaviours: “ingestion” and “other” (phase 1), and then counting the number of bites taken during each period classified as “ingestion” (phase 2). Seven dry red-pied Holstein cattle and two Blonde d’Aquitaine x Belgian White and Blue cross-breds were observed. A total of 39 h and 25 min of video were recorded and tagged for the different behaviours to train several ML algorithms. During phase 1, four different window segmentations and two different splits of the data were used to train and test four ML classification algorithms: Bagged Tree, Medium k-NN, Fine tree and linear SVM. The results show that Bagged Tree algorithms with 30 s windows and 90 % overlap gave the best results during the first phase, with an accuracy of 97.83 % for split 1 and 98.07 % for split 2. During phase 2, the same four

FAQs

What is the main goal of the study on cattle grazing behaviour?

To quantify grass-severing bite events and characterize when and how often cattle perform these bites during grazing.

How does the proposed machine learning approach work?

It uses a two-phase pipeline: first classify behaviour windows into “ingestion” vs “other”, then estimate the number of bites during windows classified as “ingestion”.

What data and sensing method are used to train and evaluate the models?

Signals come from an Inertial Measurement Unit (IMU) sensor mounted on the cows’ neck, and behaviours are supervised using video recordings tagged for different behavioural states.

Aug 5, 2026

Upload by Eden