Ear Tag Sensors Enable Data-Driven Decisions to Improve Sow Health, Reproduction, and Operation Throughput

By Marika Genzow, DVM

For decades, the swine industry has improved productivity through genetics, nutrition, housing, and management. But the next major gains are coming from a different direction: continuous, animal‑level data collected daily.

Precision livestock farming (PLF) has already transformed dairy and poultry systems, and swine is now entering its own era of data‑driven management across operations.

One of the most promising innovations being introduced is also one of the simplest: ear‑mounted sensors that continuously track behavior patterns and ear‑surface temperature in real-time.

Developed and marketed by BioCV Inc., these devices are small, durable, and engineered for commercial‑barn environments.

The true breakthrough, however, lies not in the hardware itself but in the biological signals it captures and the insights generated from those data streams.

Why? Because the system detects physiological and reproductive changes at the moment they begin altering a sow’s behavior, well before those changes become obvious to the human eye.

Across multiple research efforts evaluating the use of BioCV BioTag sensors, a consistent pattern has emerged: behavioral and temperature data from ear‑tag sensors can detect early signs of estrus, lameness, and pre‑farrowing distress (including postpartum dysgalactia syndrome).

These early signals give producers, animal caregivers, and veterinarians a chance to intervene sooner, improving sow performance and piglet outcomes.

This is not a theory. It is being validated in real barns, with real sows, under real production conditions.

WHY EAR TAG SENSORS MATTER NOW

The move toward group housing, labor shortages, and larger sow units has made it harder than ever to monitor individual animals. At the same time, the industry is under pressure to improve welfare.

Operation profitability demands reducing premature sow removals and increasing pigs weaned per sow per year.

Ear‑tag sensors offer a practical solution because they:

• monitor each sow continuously;
• require no cameras or special lighting;
• identify animals automatically;
• detect subtle behavioral changes;
• provide actionable data that integrates into existing workflows.

In short, they give producers a way to “listen” to a sow’s behavior in a way that was never possible before.

ESTRUS DETECTION: A NEW LEVEL OF ACCURACY

Estrus detection is one of the most time‑sensitive tasks in sow management today.

Missing the fertile window means lost pregnancies, more nonproductive days, and longer intervals between farrowings.

A recent production‑scale study using BioTag ear‑mounted sensors collected 1,065 labeled estrus events from 453 sows.

The research team analyzed daily activity and temperature patterns using a machine‑learning model built on 48 features, including:

• total daily activity;
• nighttime vs. daytime movement;
• posture‑change frequency;
• jerk (movement intensity);
• ear‑surface temperature;
• day‑over‑day changes in all of the above.

The breakthrough finding was that estrus is best detected by looking at how a sow changes from one day to the next.

This makes biological sense. Estrus is not defined by a single behavior but by a shift in activity structure—more walking, more posture changes, disrupted rest patterns, and slight temperature elevation observed consistently.

Using this approach, the model was able to distinguish behavioral activities correctly, identifying estrus-related behavioral shifts four out of five times. This is a strong result for a sensor‑only system operating in a commercial barn. Temperature features were also surprisingly influential, with three temperature metrics ranking among the top 10 most important.

For producers, this means:

• more accurate heat detection;
• better timing of insemination;
• fewer repeat services;
• shorter wean‑to‑service intervals;
• improved farrowing rates.

In other words, better estrus detection improves throughput— and using ear‑tag sensors make it possible at scale.

DETECTING PROBLEMS BEFORE THEY ESCALATE

Postpartum dysgalactia syndrome (PDS) is a costly disorder that affects sow health and piglet survival.

PDS is characterized by inadequate colostrum and milk production during the first days after farrowing and is caused by a variety of factors, including genetics, parity, body condition, stress, nutrition, housing, handling, climate conditions, and management practices.

The average incidence is estimated at about 13%, while reported epidemic incidence can reach 60%, causing losses of $325–$510 USD per affected sow.¹

The challenge is that the earliest signs are behavioral, not clinical. Before a sow becomes visibly sick, she often shows:

• reduced nestbuilding;
• disrupted rest patterns;
• abnormal restlessness;
• subtle temperature elevation.

These changes can appear 12–48 hours before the sow becomes febrile or off feed. Ear‑tag accelerometers within BioTags capture nestbuilding intensity and posture changes automatically, while temperature sensors detect early inflammatory shifts. When these data streams are compared to a sow’s individual baseline, deviations become clear long before a human observer notices in practice.

Research with BioTags evaluated pre‑partum activity in 148 sows.

Researchers found that sows that later developed PDS showed noticeably less nestbuilding in the days before farrowing.

In loose‑housed systems, where animal observation is challenging, this drop in activity was strong enough for an AI model to reliably flag at‑risk animals.

The model reached an AUROC of 0.812, meaning it could separate healthy and problem sows with a high level of accuracy. In crates, however, the signal was less determinant because the sow couldn’t express natural pre‑farrowing behavior.

The study demonstrated that the system could flag roughly 20% of loose‑housed sows as high‑risk with 87% accuracy.

A wider setting identifies about 40% of the herd with 67% precision and 71% recall—useful for routine monitoring and early intervention.

For veterinarians and caregivers, this is a critical insight.

Early intervention— anti‑inflammatories, hydration support, and closer monitoring—can prevent the cascade of inflammation that leads to agalactia, piglet starvation, and sow discomfort.

For producers, the benefits are certainly tangible:

• fewer stillbirths;
• improved colostrum production and piglet intake;
• reduced piglet mortality;
• fewer sow removals;
• smoother farrowing‑room flow.

This is where PLF enabled by BioTags, combined with the machine‑learning capabilities of the platform from BioCV, delivers its greatest value: preventing problems before they become crises.

LAMENESS: ANOTHER CASE STUDY IN EARLY INTERVENTION

Lameness is another condition where early signs are subtle and late signs are costly.

In the Production Analysis Summary for the U.S. Pork Industry, a 2024 MetaFarms analysis found that 17.9 percent of sows that died or were euthanized were lost due to locomotion issues—including lameness, broken bones, feet and leg injuries, and spraddles. Nearly 22 percent of all deaths occurred in first‑parity animals.²

With the investment in a replacement gilt approaching $500 per head, timely detection and intervention before lameness becomes apparent and severe is an economic must‑do.

Lameness research using BioTags helps demonstrate that these sensors detect:

• reduced movement;
• altered day–night activity;
• temperature shifts associated with inflammation;
• changes in head‑movement patterns.

Again, the key insight gained through initial research shows that individual animal baselines outperform herd averages as an indicator of lameness. Sow‑to‑sow variation was so large that population averages did not separate sound from lame in a useful way.

Using individualized baselines established for each sow in as little as two weeks, the model achieved a recall of 0.76 at a high-recall threshold (0.28) and a precision of 0.56. around 0.90 at a high‑sensitivity operating point—meaning half of the alerts generated were false positives—but note that this is not a flaw. It’s a tradeoff.

A false positive costs a quick check by the caregiver. Without timely treatment, a missed true positive may become a severe case without timely treatment.

This is the same logic that applies to estrus and PDS detection: early always provides a better, more economical outcome than late.

THIS TECHNOLOGY IS THE NEXT STEP IN PRECISION LIVESTOCK FARMING

Across estrus, PDS, and lameness, the same biological principle holds. Health and reproductive events change how a sow behaves before they change how she looks.

Ear‑tag sensors capture: activity structure; posture transitions; circadian rhythm; restlessness; and temperature fluctuations.

And because they operate continuously, they detect deviations that humans often miss, since caregivers are typically present only eight of 24 hours on most commercial farms—perhaps fewer on weekends.

The use of BioCV BioTags requires no infrastructure beyond the tags and gateway receivers, operates continuously in any lighting condition, and provides individual animal identification by default.

This makes them uniquely suited for commercial barns, where camera systems may be impractical.

WHAT THIS MEANS FOR THE FUTURE OF SWINE PRODUCTION

The swine industry is entering a new phase of precision livestock farming—one where individual animal data becomes the foundation of herd management moving forward.

Ear‑tag sensors are leading this shift because they are: affordable; easy to deploy; validated in real barns; biologically meaningful; and actionable for staff and veterinarians.

They don’t replace stock people or change the biology of a sow—they simply help us see problems earlier, prevent disease from escalating, and make every intervention more effective across farms.

For producers looking to improve throughput, reduce sow removals, and increase pigs weaned per sow per year, this technology represents a practical, validated step forward.

And for veterinarians, it offers a new tool for early detection, targeted intervention, and improved welfare outcomes in practice.

This is the next innovation in precision livestock farming—and it’s already here. For more about this cutting‑edge IoT technology and real‑time animal health monitoring, estrus detection, and farrowing prediction for your operation, visit www.biocv.info/lisa-sow.

References

¹ Kemper N. Update on postpartum dysgalactia syndrome in sows. J Anim Sci. 2020 Aug 18;98(Suppl 1): S117-S125. doi: 10.1093/jas/skaa135. PMID: 32810252; PMCID: PMC7433910.

² MetaFarms; National Pork Board. Production Analysis Summary for U.S. Pork Industry: 2020–2024. Pork Checkoff. https://porkcheckoff.org/research/production-analysis-summary-for-u-s-pork-industry-2020-2024/

Marika Genzow, DVM
DipECPHM

Marika Genzow, DVM, is a diplomate of the European College of Porcine Health Management. She spent more than 30 years as a global senior technical service veterinarian for Boehringer Ingelheim, contributing to swine disease management worldwide. Her work includes advancing early respiratory‑disease detection through precision livestock farming and sensor‑data analysis.