triact package for R: analyzing the lying behavior of cows from accelerometer data.

Area of Science:

• Animal Science
• Agricultural Technology
• Bioacoustics and Biomechanics

Background:

• Accelerometers are effective sensors for monitoring animal behavior, particularly the lying patterns of cows.
• Researchers often require custom data analysis scripts for algorithm transparency and control, preferring open-source solutions over proprietary software.

Purpose of the Study:

• To introduce the 'triact' R package, an open-source tool designed for analyzing cow lying behavior.
• To provide animal scientists with a user-friendly workflow for processing raw accelerometer data.
• To enable the calculation of key lying behavior metrics and activity proxies.

Main Methods:

• Development of the 'triact' R package, compatible with manufacturer-agnostic triaxial accelerometer data.
• Implementation of algorithms for determining total lying duration, number of lying bouts, and mean duration of lying bouts.
• Inclusion of functionalities for assessing lying laterality and calculating proxies for physical activity levels.

Main Results:

• The 'triact' R package offers a streamlined workflow for analyzing cow lying behavior from accelerometer data.
• The package facilitates the quantification of essential lying metrics, including duration, bout frequency, and bout length.
• Provides capabilities for analyzing lying side preference and estimating general cow activity.

Conclusions:

• The 'triact' R package is a valuable, open-source resource for researchers studying cow behavior.
• Its user-friendly interface and comprehensive functionalities enhance the transparency and efficiency of data analysis.
• Freely available via CRAN, promoting wider adoption in animal science research.