Tjotta accelerometer monitored lambing dataset.

Area of Science:

• Animal Science
• Agricultural Engineering
• Biomedical Engineering

Background:

• Early lambing detection is vital for sheep welfare and farm management.
• Wearable inertial sensors show promise for monitoring sheep behavior, particularly transitions between standing and lying.
• Existing real-time accelerometry systems for small ruminant lambing detection are lacking.

Purpose of the Study:

• To establish a dataset for developing real-time automated lambing detection in sheep.
• To identify key behavioral indicators of lambing using accelerometry data.
• To provide a foundation for future algorithms classifying birth-related behaviors.

Main Methods:

• Monitored 61 ewes using accelerometers at 20 Hz.
• Simultaneously recorded lambing events via video for precise timing (113 events).
• Collected supplementary data: litter size, need for assistance, birth year, lambing history, and ultrasound results.

Main Results:

• Collected comprehensive sensor and video data from 113 lambing events.
• Established precise birth times and documented associated behaviors.
• Created a dataset linking accelerometer data to lambing events for algorithm development.

Conclusions:

• The developed dataset is a valuable resource for creating automated lambing detection systems for sheep.
• This research supports the advancement of precision agriculture through improved animal monitoring.
• Further development of algorithms can enhance sheep welfare and farm management efficiency.