Slogans such as “Rise Early to Farm Pigs, Call AI to Help!” and “Excel in Intelligent Pig Farming, Marry a Pretty Wife Early!” are appearing on walls across China’s southern countryside provinces. The campaign is part of the Sichuan pig farming corporation Dekon Group and pig feed supplier Tequ Group’s new partnership with Alibaba Cloud to apply its AI-powered “ET Brain” to pig farming. The trio are investing tens of millions of USD in the project, which was announced on February 6, 2018.
Over the past year, Alibaba has implanted its ET Brain in the aerospace, transportation, environment, and healthcare sectors, fast-tracking China’s social infrastructure revolution. Recent food contamination scandals have made food safety a pressing issue in China, and the agricultural industry a next logical application for ET Brain.
China’s pork production accounts for more than half of the world supply, while its per capita pork consumption ranks 3rd. By 2020, Tequ Group sales will exceed 10 million tons, while Dekon will breed up to 10 million pigs annually. This is an opportunity for artificial intelligence to optimize operations, and both companies are actively building their IoT and Enterprise Resource Planning (EPR) systems.
On pig farms each pig wears a wireless radio-frequency identification (RFID) tag. These are pricey and difficult to scan, making and farmers must individually log data into mobile applications or fill in paper forms.
This is where computer vision and voice recognition AI can help. Real-time video footage is collected through surveillance cameras. Using computer vision, ET Brain will set up profiles for each pig, documenting their breed, age, weight, eating conditions, exercise intensity and frequency, and movement trajectory. The first phase of the launch includes functions such as herd behavior analysis, inventory count, health monitor, and automatic weighing.
One challenge is telling pigs apart — Alibaba considered applying facial identification to pigs to no avail. Instead they tattooed identifying numbers on the pigs’ bodies.
ET Brain tracks the entire production line. Based on behavior tracking, gilts are selected for mating. After they give birth to piglets, usually in a litter of ten, ET Brain will use voice recognition to ensure the little ones are not suffocated by their mothers’ weight. This lowers death rate by 3%, increasing annual production rate by three piglets per sow.
Aside from breeding, feeding and weighing, other important steps in pig farming are disease control and epidemic monitoring. ET Brain will analyze pigs’ behavior, acoustic characteristics and infrared temperature measurements, to determine the health status of pigs, targeting epidemic early warning signs and specialized vaccinations.
Alibaba Cloud has dispatched a team of algorithm engineers, product developers, and video analytics team to Sichuan work on the project, while Tequ will add experts on pig farming.
“Our core solution is to reduce the reliance on farmers and dependence on equipment through automated video analytics,” explains Alibaba Cloud’s Sheng Zhang, who added that the use case is highly replicable.
Alibaba’s AI solutions have thus far been more widely deployed in urban environments. Its “ET City Brain” focuses on improving China’s urban infrastructure using capabilities such as voice, image, text recognition, and natural language processing. Earlier last month, the company announced it will deploy ET City Brain in the Malaysian capital of Kuala Lumpur.
However, it is also worth considering the appropriateness of AI applications for lesser developed areas. Today, 43% of the Chinese population lives in the countryside. Granting this rural economy access to AI is a difficult but important task, especially for underdeveloped industries like pig farming that have limited access to new technology.
Journalist: Meghan Han | Editor: Michael Sarazen