Livestock health sensors: small signals, big decisions

What the sensors actually measure

Livestock health sensors fall into a few groups. Wearables on the animal, such as neck collars, ear tags, leg bands and pedometers, use accelerometers to measure movement, lying, eating and rumination. Rumen boluses sit in the reticulum and record temperature and sometimes activity. Fixed systems in the building, such as cameras and microphones, watch mobility, body condition, behaviour or coughing across a whole pen. Weighing systems, from walk-over platforms to automatic crates, track growth and flag animals falling behind.

None of these devices diagnoses disease. They detect change: a cow ruminating less than her normal pattern, a heifer showing a spike in activity, an increase in coughing in a calf shed, a lamb that has stopped gaining weight. The value lies in noticing that change a day or two earlier than a stockperson could, and in doing it every hour for every animal.

How accurate are they?

A 2021 review in the journal Animals by researchers at Chungnam National University pooled validation studies of wearable sensors for cattle. It found very strong agreement between sensors and direct observation for lying time, and strong agreement for eating and rumination time. Drinking and walking were much weaker: sensitivity for drinking was about 22% and for walking about 34%, although specificity for walking was high.

The same review found considerable differences in performance between studies, including between studies of the same brand, and noted that most papers did not report enough detail to explain why. It also listed practical limits: battery life from around 60 days to 10 years depending on the device type, wireless range from about 75 m to 3 km, and algorithms that manufacturers keep confidential.

The practical takeaway is to trust sensors most for the behaviours they measure well, such as rumination and lying, and to be sceptical of claims built on measurements with weaker evidence. Ask suppliers for independent validation data, not just their own figures.

Alerts are only useful if someone acts

Penn State Extension's guidance on precision dairy technologies makes three points that apply to any livestock system. First, these tools offer precision rather than diagnostic accuracy, so an alert is a prompt to look at the animal, not a verdict. Second, alert thresholds usually need adjusting to each farm's environment to reduce false alerts; out-of-the-box settings can bury staff in warnings. Third, labour is reallocated rather than removed: someone must maintain the equipment, check alerts every day and coordinate responses such as breeding, treatment or a vet visit. Without that commitment, and veterinary input during set-up, adoption tends to fail.

Before buying, agree an alert routine in writing.

• Who checks the alert list, at what times of day, and who covers days off?
• What does each alert type trigger: observe, take a temperature, call the vet, or record only?
• How will you review false alarms after the first month and adjust thresholds?
• How will treatments and outcomes be recorded so you can tell whether the system is catching problems earlier?

Matching the technology to the livestock system

Housed dairy herds are the most common users of wearables, because the animals pass a reader or gateway several times a day and fertility and health events are costly. Beef and sheep systems, especially at grass, often get more value from weighing and electronic identification: an EID reader linked to a weigh crate identifies animals that are not growing, which is often the earliest practical sign of worms, lameness or poor grazing. In pig and poultry buildings, cameras and microphones that watch the whole group can be more practical than tagging individuals.

Connectivity decides a lot. Check where gateways need to go, whether there is mains power and signal at each site, and how data reaches your phone when animals are in outlying fields.

Grants in England

The Farming Equipment and Technology Fund 2026 animal health and welfare theme included several monitoring items. The automatic health and welfare remote monitoring system (FETF315CA) had an expected average cost of £17,549 and a grant of £8,774. It had to monitor at least one health and welfare parameter, such as activity, feeding behaviour or body temperature, using ear tags, collars, leg bands or boluses, and trigger alerts. Systems that only detect heat, tail-mounted devices, devices applying an electrical pulse, and Cattle Tracing System tags were excluded, and ongoing subscriptions were not eligible. The expected cost was based on one base unit and around 200 animal devices.

Other items included a visual or acoustic monitoring system for cattle housing (FETF613, grant £4,180), an automated mobility or body condition scoring camera (FETF235A, grant £3,253), an automatic cattle weighing unit (FETF67, grant £2,427) and an automatic weighing and drafting crate for sheep (FETF72, grant £7,996). FETF 2026 closed on 12 May 2026, and Defra has said future grants will be combined into a new scheme from 2027.

A buyer's checklist

These checks apply whether you are buying 50 ear tags or a whole-herd collar system.

• Battery life in real use, replacement cost per device, and whether batteries or whole units are replaced.
• Independent validation for the specific measurements you care about, ideally in similar breeds and systems.
• Data export: can you get raw or summary data out in a standard format, and share it with your vet?
• Integration with your existing herd or flock software and EID equipment.
• Full five-year cost including subscriptions, gateways, replacement devices and installation.
• How the supplier handles lost or damaged devices and firmware updates.

Measuring the payback

The fairest test is to record what happens now, then compare after the system goes in. Useful measures include days from calving to first service, conception rate, number of clinical cases caught early versus late, treatment costs, deaths and culls, and staff time spent observing animals. Give it at least a full season before judging. If you also run handling equipment, loaders or feeders around the sensors, use FarmFleets search to check that yard machinery fits the building layout; for dairy-specific automation see our dairy automation guide.

Sources

Sources checked 2026-09-25

• Lee and Seo, Wearable Wireless Biosensor Technology for Monitoring Cattle: A Review, Animals (2021), via PubMed Central — checked 2026-09-25
• Penn State Extension, Precision Livestock Farming: Dairy Technologies — checked 2026-09-25
• GOV.UK, Animal health and welfare items and specifications, FETF 2026 — checked 2026-09-25
• GOV.UK, Apply for the Farming Equipment and Technology Fund (FETF) 2026 (closed) — checked 2026-09-25

Corrections and feedback

See our Corrections Policy, contact the team, or Report incorrect information.