Livestock behaviour analysis using high-resolution positioning
Long-duration deployment of terrestrial positioning architecture enabling continuous observation and quantitative analysis of animal movement, behaviour, and environmental response within operational livestock environments.
Context
Traditional observation methods are limited in temporal resolution and scalability, making it difficult to capture continuous behavioural patterns or detect subtle changes over time.
Understanding animal behaviour within operational livestock systems is critical for improving welfare, productivity, and environmental management.
High-resolution positioning enables continuous observation and quantification of behaviour within real-world environments.
Deployment of positioning infrastructure within operational livestock environments.
Deployment Overview
The positioning architecture was deployed within operational dairy environments to enable continuous tracking of animal movement over extended periods.
Deployments supported continuous tracking of over 80–100 animals simultaneously, generating datasets comprising tens to hundreds of millions of validated position measurements.
Omnisense provided the positioning infrastructure and data foundation, ensuring spatial data was robust, continuous, and suitable for downstream behavioural modelling and analysis conducted by research partners.
The system utilised distributed terrestrial nodes to provide high-resolution positioning within indoor barn environments, capturing individual and group behaviour without reliance on GNSS.
Deployment duration and data continuity enabled analysis of behaviour over time rather than isolated observations.
Individual movement trajectory showing transitions between functional zones (feeding, resting, milking) within a livestock environment.
Data Collection and Methodology
Positioning data was collected continuously at high temporal resolution, enabling reconstruction of detailed individual movement trajectories.
Datasets included continuous x,y trajectories per animal, enabling derivation of spatial and behavioural metrics including zone occupancy, inter-animal distance, proximity to features of interest, and transition frequency between functional zones.
Data processing enabled trajectory reconstruction, spatial occupancy analysis, clustering and grouping behaviour, and temporal activity patterns.
The resulting datasets support both descriptive behavioural analysis and quantitative research investigations.
Behavioural insights
Continuous positioning data enables direct observation and quantification of behavioural patterns under varying environmental and operational conditions.
Space Use and Social Behaviour
Long-duration data supports analysis of space utilisation and interaction between individuals.
Indicators include reduced movement frequency, altered walking behaviour, and increased stationary periods.
Activity and Lameness Indicators
Continuous tracking enables detection of subtle changes in locomotion and activity patterns.
Indicators include reduced movement frequency, altered walking behaviour, and increased stationary periods.
Heat Stress Behaviour
High-resolution positioning data enables identification of clustering and bunching behaviour during elevated temperature and humidity conditions.
Observed responses include redistribution toward feeding zones, reduced cubicle transitions, and increased spatial clustering under thermal load.
Example Data Outputs
The following examples illustrate outputs derived from continuous positioning data across individual and group scales.
These outputs enable quantitative behavioural analysis at a resolution not achievable through conventional observation or sensor-based approaches.
Individual Trajectory
Individual trajectory extracted from high-density multi-animal dataset, illustrating behavioural patterns within shared space.
Positioning Accuracy (CEP)
Positioning accuracy characterised through CEP analysis, indicating spatial variation in ranging performance.
Multi-Animal Movement
Multi-animal movement trajectories illustrating spatial distribution and interaction patterns within a livestock environment.
Spatial Occupancy
Spatial occupancy heatmap derived from positioning data, indicating preferred zones and activity intensity over time.
Research Collaboration
Deployments were conducted in collaboration with academic partners including the University of Reading and Writtle University College as part of structured research programmes investigating animal behaviour and environmental response.
The positioning system formed part of the experimental infrastructure, with datasets generated to support independent statistical analysis, modelling, and publication.
Outcomes
The deployment demonstrates reliable operation of the positioning architecture within complex real-world environments while generating datasets of sufficient quality and duration to support quantitative behavioural research.
The system operates as a persistent measurement layer within the research environment, enabling continuous spatial data capture aligned with experimental study design.
This establishes a foundation for broader application of positioning-enabled behavioural analysis across livestock systems and related domains.
Further details, including peer-reviewed publications and technical outputs arising from these deployments, are available on the Publications page.