Focusing on making pig raising simpler, Muyuan tried to build a large model of pig technology

Top news reporter Yang Guifang

On June 22, 2024, the 19th Muyuan Pig Festival of Muyuan Group was held in Nanyang, where Muyuan headquarters is located. Although it seems to be an internal commendation conference, the event has become a technical event in the aquaculture industry through 19 consecutive sessions.

At the event site, the top news reporter noticed that the complete cost of some of the field lines in Muyuan has dropped to less than 13 yuan per kilogram.

【Breakthrough】The complete cost of pig breeding in some farms has been as low as 13 yuan/kg

In the commendation meeting on the same day, three awards were set up: the Performance Achievement Award, the Excellent Batch Award & the Leading Batch Award and the Innovation Achievement Award. Among them, in terms of excellent batch awards & leading batch awards, the reporter noticed that the cost of pigs in a batch of farms has dropped to below 13 yuan/kg, such as the complete cost of pig breeding in 12 lines in Huaxian County, northern Henan Province reached 12.8 yuan/kg in May. The complete cost of pig breeding in the first 6 fields in Anhui reached 12.5 yuan/kg.

In particular, it is worth mentioning that in the past, the performance of some high-cost field lines has been rapidly improving, and the cost has dropped to 13-13.5 yuan/kg, and even some field lines have reached the level below 13 yuan/kg.

This means that under Muyuan's plan to continue to tap the potential of 600 yuan, the company has a number of field lines to achieve rapid improvement in performance and continuous exploration of cost space. Excellent field lines are constantly breaking through the upper limit, the rapid reversal of high-cost field lines, and the dispersion of breeding costs are rapidly tightening.

Not long ago, Muyuan Co., Ltd. said on the investor interactive platform that the company's complete cost of pig breeding in May 2024 is about 14.3 yuan/kg, which is 0.7 yuan/kg lower than the complete cost of pig breeding in 2023. According to the previous public information, the company is confident that the complete cost of pig breeding will be reduced to 13 yuan/kg by the end of the year. Judging from the current performance, it is foreseeable that with the steady improvement of the performance of each field in the next few months, the realization of its cost target has strong certainty.

Gao Tong, chief financial officer of Muyuan, mentioned in his work report that there is still a certain gap from the world's leading and excellent level. "Seeing the international advanced has given us stronger motivation and higher work standards, and our leading regional field line should launch a charge towards such a goal."

Algorithmic models are influencing the future of animal husbandry

Behind the above data is the support of technology from the "meta point" and business from the "bottom breakthrough". For example, at the commendation meeting, a number of young scientific and technological personnel who graduated from colleges and universities and devoted themselves to animal husbandry were commended. This also reflects that in Muyuan, the power of innovation and technology has emerged, and the future pig breeding industry will be a competition of science and technology.

What do the different sounds made by a pig represent? What is the difference in sound between healthy and abnormal pigs? How can the sound of each pig on the farm be fully monitored? These require a large amount of data collection, and repeatedly check with veterinary personnel, constantly proofread standards, formulate results, and finally form the base of the sound early warning system algorithm.

Wang Liusong, a young man born in the 90s, is one of the technicians of the veterinary intelligence project-AI sound warning project. In order to improve the effect of AI early warning and achieve the goal of "listening to sound and distinguishing disease", Wang Liusong's team did a lot of data collection, and in order to adapt the equipment to different environmental conditions, it drove 2,000 kilometers from Nanyang to Gansu, and drove for two days and two nights to the local Muyuan farm for a week of data collection.

It is reported that at present, Muyuan's sound monitoring equipment has gone through 3 iterations, and in October 2023, each line of Muyuan Farm has been covered, and the current system's early warning accuracy for pig cough has been quite accurate.

Some people are obsessed with building pig herd sound algorithms, and some people are obsessed with building picture recognition schemes.

Disease discrimination through a picture not only tests the inclusion of a large number of image data into the algorithm library, but also tests whether the data standard for disease discrimination can be established. For example, whether there is enough image data to support the data base and transfer people's medical cognition to models and algorithms.

Nothing is a fantasy, as long as you have a dream and put it into practice to subvert it, you may reap a little goal. It's like when the commended team is introduced at the Pig Festival, they all have an identity: from 0-1 to the leading big brother, subversive, dreamer.

The reporter noticed that at this commendation meeting, a number of algorithm engineers were rewarded, such as some engineers who focused on pig behavior recognition, some engineers who focused on opening up pig body temperature monitoring, and some engineers who focused on establishing intelligent diagnostic models and platforms.

An algorithm engineer told reporters: "In the future, we hope to make pig raising simpler and make accurate judgments without relying on human experience." ”

【Exploration】From point to line, we hope to build a large model of pig science and technology

The development history of China's thousand-year-old aquaculture industry reflects the iterative history of productivity from agriculture to industrialization and then to intelligence.

In the era of All in AI, every economic entity is looking for and exploring ways and means of AI empowerment, and the AI wave is also intensifying China's agriculture and animal husbandry to introspect, innovate and make breakthroughs in intelligence again.

At this Muyuan Pig Festival, the issue of innovation was also highlighted. The theme of the conference, "Vision Leads to Create the Future, Innovation Drives to Break the Ceiling", inspires young people who devote themselves to science and technology, agriculture and animal husbandry.

According to the data, 40,000 of Muyuan's staff are university graduates, and nearly 7,000 are engaged in intelligent research and development and application. From 2018 to 2023, Muyuan's pig slaughter volume has increased by about 4.8 times, and the full-caliber innovation and R&D investment has exceeded 30 billion yuan.

The directors of the farm are full of imagination about scientific and technological pig breeding. Pang Kang, director of the line growth farm in Dali 1 Field 3 in Shaanxi Province, used "shock" to describe the changes brought by intelligence and big data to the breeding industry, because he can now view all the breeding units through his mobile phone even if he is resting in bed, and abnormal information such as pig fever and cough can be displayed on the mobile phone at any time. "Intelligence is already a trend, and it will only be more intelligent in the future, and it will be easier for employees to raise pigs." Pang Kang said.

Fan Junzhong, director of the 21st line breeding farm in Muyuan Neixiang, believes that with the trend of intelligence and the Internet of Things and AI empowering thousands of industries, the future breeding industry is a more technological industry, and intelligence will help pig farmers, find problems in a more timely manner, and manage more simply, and the breeding efficiency will be higher. He is convinced: "I believe that the future will definitely be realized." ”

Fu Xin, director of the 21st line growth farm in Muyuan Neixiang, believes that "the future is the world of science and technology, and pig breeding can be more avant-garde." ”

These wishes of the field managers have been written on the work vision list of scientific and technological research and development personnel. Zhao Zhiyuan, the person in charge of the AI sound warning project, hopes to move up step by step from "veterinary intelligence" to achieve "intelligent pig raising".

"At the bottom of the floor, we explored a point, and then how to explore a line." Jin Pengtan, the person in charge of the Muyuan needle-free injection project, mentioned that he hopes to be more confident when he is benchmarked against foreign scientific and technological breeding technology, "We will study, but we will not follow their path, because we must realize our own technical system." ”