China's mountainous province embraces AI in agriculture

GUIYANG -- At a Silkie chicken breeding base in Tiantaishan village in the city of Chishui, Southwest China's Guizhou province, each coop is equipped with a real-time surveillance camera and each chicken fitted with a microdevice.

Such technology enables breeders to track the chickens' daily steps and identify the sick ones. Additionally, sensors installed along the coop fences can monitor predator threats and detect any structural damage.

"Artificial intelligence (AI) has significantly improved the breeding standard of our Silkie chickens," said Xu Qiyong, who manages the breeding base.

According to Xu, healthy Silkie chickens typically take 10,000 to 20,000 steps a day. If the system detects a chicken walking fewer than 1,000 steps, it signals the possibility of illness and the staff will intervene promptly.

As digital technologies rapidly advance, AI has empowered farming with data-driven precision and increased efficiency in Guizhou, where 92.5 percent of the land is mountainous or hilly.

In April 2023, a team formed by Tencent Cloud and the Tencent Class of Shenzhen University jointly implemented a smart Silkie chicken breeding system at the base, leveraging AI technology and cloud platforms.

Thanks to this system, the team alleviated issues such as disease control and wild animal invasions -- boosting the base's productivity by 30 percent within a short period.

Additionally, Tencent Cloud and the Tencent Class of Shenzhen University are collaborating with Guizhou University and other institutions to develop a large language model for the poultry industry.

This AI-driven system will serve as a virtual assistant to help breeders tackle challenges faced in the Silkie chicken farming.

"Since last year, we have sold 100,000 Silkie chickens and 300,000 eggs, and demand still exceeds supply," Xu noted, adding that he is working with the local government to encourage more farmers to adopt AI-driven chicken farming.

This shift from traditional experience-based farming to data-driven methods in poultry farming is not an isolated case. In Guizhou's Caohai Township, Zhongken Potato Industry Co Ltd has developed a "5G + digital agriculture" platform to digitize and automate potato breeding processes.

At the company's digital potato breeding base, temperatures and humidity levels in the greenhouses are monitored in real-time, while the irrigation system is automatically controlled.

Moreover, infrared sensors at the base can detect potential pest infestations, reduce the risk of large-scale crop damage, and ensure stable seedling growth in a controlled environment.

"We have established a potato seed breeding center to produce premium potato seeds. In the past, farmers could harvest at most one tonne of potatoes per mu (about 0.07 hectares). With the improved seed varieties we have developed, the yield now has increased to two to three tonnes per mu," said Li Riyu, general manager of Zhongken Potato Industry Co Ltd.