(Yicai Global) Feb. 7 -- Alibaba Cloud, Sichuan Tequ Group, and DEKON Group will cooperate in targeted training and research and development of the former's ET Brain to fully realize AI-based pig breeding, they announced.
Investment in this project will total hundreds of millions of yuan. During its early-phase theory verification, ET Brain helped increase sows' annual reproductivity by 3 piglets and lowered the death and attrition rate by about 3 percent, Leiphone.com reported.
ET Brain is a super-intelligent technology Alibaba Cloud introduced in December. It integrates comprehensive capabilities, such as city management, industrial optimization, complementary medicine, environmental amelioration and air-traffic control, and thus lends itself to swift, optimal real-time decision-making in complex situations requiring multidimensional perception, comprehensive insights and continuous evolution.
Pig farms have much historical video and sow reproduction data that can help ET Brain learn in-depth sows' reproduction models, Alibaba Cloud said.
Based on machine vision technology, ET Brain creates a file for each pig detailing the breed, age (in days), weight, feeding, exercise intensity and frequency, and thereby analyzes pigs' behavioral and feeding traits and feed to weight-gain ratios. Together with acoustic features and infrared temperature measurements, ET Brain can also tell whether a pig is sick (e.g. by detecting coughing) and issue illness alerts.
ET Brain also predicts the annual reproductivity of each sow -- the number of live weaned piglets she produces each year -- an important indicator for measuring the economic benefits of a pig farm and a sow's reproductive performance. ET Brain will suggest culling sows that demonstrate declining reproduction.
Tequ Group, Alibaba Cloud's partner, is China's leading company in the field of agricultural industrialization. Tequ Group is expected to achieve annual feed sales of more than 10 million tons, with the volume of its annual pig feeding and poultry farming surpassing 10 million and 250 million, respectively, by 2020.
JD.Com Inc. has also moved into animal husbandry, setting up its AI chicken project last year. Liu Qiangdong posted the 'running chicken' project, JD.Com's anti-poverty program, on social media in May. JD.Com requires that 'running chickens' be free-range, with step counters attached to their feet. JD.Com will buy any chicken that runs over one million steps for three times the going rate.
JD.Com will integrate its Internet of Things system into the chicken farming process, thus realizing the connection between the production quantitative model, breeding management system and shelter environmental control system that applies an AI algorithm. In addition to automatic feeding, watering and waste removal, the system will also monitor and identify a chicken's food intake, defecation and other physiological conditions, and diagnose diseases by distinguishing different clucks. If a chicken is sick, epidemic experts provide medical treatment and prescribe drugs online.