Proximity Sensors Calibrated to Track Sheep Social Contacts Accurately

Proximity sensors that log contact between animals could transform how researchers study livestock behaviour, welfare and disease spread, but only if their signals can be trusted. A new study published in Animal Biotelemetry reports that commercially available sensors based on received signal strength indicator (RSSI) can reliably detect biologically meaningful social interactions in sheep, provided they are carefully calibrated for the environment in which they are used.

Understanding social contacts in livestock is essential for behavioural research, welfare assessment and modelling pathogen transmission. RSSI-based proximity logging is a promising tool for capturing fine-scale interactions in precision livestock farming, yet sensor performance depends strongly on environmental conditions and device configuration. The research team, led by Edoardo Fiorilla and colleagues, set out to calibrate and validate these sensors under controlled indoor and outdoor conditions and then test them in live sheep.

The calibration experiments covered inter-individual distances from 0.5 to 5 metres. They showed a predictable decrease in RSSI as distance increased, but the pattern was modulated by the environmental setting and by the orientation of the sensors, both of which significantly influenced signal attenuation. In practical terms, the same pair of devices could report different signal strengths depending on whether animals were indoors or outdoors, or how the tags were positioned on their bodies.

To turn raw signal values into usable distance categories, the researchers applied probability-based modelling. This demonstrated that RSSI reliably predicts biologically relevant proximity thresholds of 0.5 metres, 1.0 metre and 1.5 metres or less. Importantly, the results support deriving context-specific RSSI cut-offs for each deployment rather than relying on universal thresholds that would hold everywhere.

The decisive test came from in vivo validation with sheep. The team combined proximity sensor data with image-based distance measurements matched to each individual RSSI detection event, allowing them to check how often the sensors classified a contact into the correct distance class. The overall classification accuracy was 82 percent.

Performance varied by distance class. Per-class F1-scores ranged from 0.76 for distances greater than 1.5 metres to 0.86 for distances of 1 metre or less, indicating that the sensors were somewhat better at recognising close contacts than at ruling out distant ones. Per-class F1-scores ranged from 0.76 for distances greater than 1.5 metres to 0.86 for distances of 1 metre or less.

When appropriately calibrated, RSSI-based proximity sensors can reliably detect biologically meaningful social interactions in sheep across a range of inter-individual distances. The findings support proximity sensing as a precision livestock farming tool for studying social behaviour, welfare assessment, management and contact-based disease processes, but the emphasis on calibration makes clear that the technology is not plug-and-play.

For researchers considering RSSI loggers, the practical message is to run context-specific calibration before deployment, account for orientation effects, and validate sensor classifications against an independent measure such as image-based distance estimation.

Subject of Research: Calibration and validation of RSSI-based proximity sensors for measuring social contact distances in sheep

Article Title: How close is close enough? Calibrating and validating RSSI-based proximity sensors for livestock social interaction studies

Article References: Fiorilla, E., Pittarello, M., Bordin, C., Cianchetti, M., Nurisso, S., Perona, G., Schiavone, A., & Ozella, L. (2026). How close is close enough? Calibrating and validating RSSI-based proximity sensors for livestock social interaction studies. Animal Biotelemetry. https://doi.org/10.1186/s40317-026-00492-y

Image Credits: AI Generated

DOI: 10.1186/s40317-026-00492-y

Keywords: proximity sensors, RSSI, livestock, sheep, precision livestock farming, social behaviour, animal biotelemetry, pathogen transmission, calibration, welfare assessment, close, enough