A revolutionary new technology that uses artificial intelligence (AI) to determine whether pigs are pregnant has been developed. It can potentially cut annual veterinary costs by about $8.8 million.
On August 7th, the Rural Development Administration (RDA) unveiled this innovative AI model. The technology employs an algorithm that analyzes ultrasound images to confirm pregnancy status. The RDA has enhanced low-quality ultrasound images and trained the AI model to create a user-friendly pregnancy diagnosis tool for farmers.
Traditionally, farmers have used estrus detection and ultrasound to determine pregnancy. This involves monitoring pigs based on a 21-day estrous cycle or having experts interpret ultrasound images after 25 days. This process can be challenging for farmers due to its reliance on specialized knowledge.
The new AI-driven process simplifies pregnancy detection for farmers. By capturing abdominal ultrasound images of sows for just over 10 seconds, the AI analyzes the data and provides a result. The program has demonstrated over 95% accuracy in diagnosing pregnancy between 22 and 25 days post artificial insemination.
Importantly, if the ultrasound equipment can store images, farmers can use this AI program without incurring significant additional costs. The RDA emphasized that this technology makes pregnancy detection accessible to non-professionals.
The benefits extend beyond diagnosis. Farmers can now extend the health care period for pigs, leading to increased productivity. Additionally, identifying non-pregnant sows allows for timely re-insemination, reducing non-productive days and saving on feed costs. For a breeding stock of 900,000 sows, this program could save around $2.5 million in labor costs and approximately $6.4 million in feeding costs annually.
The RDA has filed three patent applications related to AI technology, including those for deep learning models for pregnancy diagnosis, data generation devices, and diagnostic methods. The administration plans to transfer these technologies to relevant industries.
Kim Im Ki, director of the National Institute of Animal Science at the RDA, stated, “In addition to pregnancy determination, we are developing various AI technologies to manage sow health and detect piglet risks. Our goal is to enhance the competitiveness of pig farms by expanding smart farming technologies that increase productivity while reducing breeding costs.”
Most Commented