Login / Signup

Proximity tracking using ultra-wideband technology for equine social behaviour research.

Laura Torres BordaPeter M RothJennifer LumetzbergerUlrike AuerFlorien Jenner
Published in: Scientific reports (2024)
Sociopositive interactions with conspecifics are essential for equine welfare and quality of life. This study aimed to validate the use of wearable ultra-wideband (UWB) technology to quantify the spatial relationships and dynamics of social behaviour in horses by continuous (1/s) measurement of interindividual distances. After testing the UWB devices' spatiotemporal accuracy in a static environment, the UWB measurement validity, feasibility and utility under dynamic field conditions was assessed in a group of 8 horses. Comparison of the proximity measurements with video surveillance data established the measurement accuracy and validity (r = 0.83, p < 0.0001) of the UWB technology. The utility for social behaviour research was demonstrated by the excellent accordance of affiliative relationships (preferred partners) identified using UWB with video observations. The horses remained a median of 5.82 m (95% CI 5.13-6.41 m) apart from each other and spent 20% (median, 95% CI 14-26%) of their time in a distance ≤ 3 m to their preferred partner. The proximity measurements and corresponding speed calculation allowed the identification of affiliative versus agonistic approaches based on differences in the approach speed and the distance and duration of the resulting proximity. Affiliative approaches were statistically significantly slower (median: 1.57 km/h, 95% CI 1.26-1.92 km/h, p = 0.0394) and resulted in greater proximity (median: 36.75 cm, 95% CI 19.5-62 cm, p = 0.0003) to the approached horse than agonistic approaches (median: 3.04 km/h, 95% CI 2.16-3.74 km/h, median proximity: 243 cm, 95% CI 130-319 cm), which caused an immediate retreat of the approached horse at a significantly greater speed (median: 3.77 km/h, 95% CI 3.52-5.85 km/h, p < 0.0001) than the approach.
Keyphrases
  • healthcare
  • mental health
  • high resolution
  • machine learning
  • blood pressure
  • mass spectrometry
  • electronic health record
  • heart rate
  • bioinformatics analysis