When e-commerce giants and technology companies have pressed on agriculture, it reflects the next trend of Internet competition-the heavier model, entering the upstream of the industry.
Editor’s note: This article is from the micro-channel public number “geek park” (ID: geekpark) , Author: Vivi.
When e-commerce giants and technology companies have pressed on agriculture, it reflects the next trend of Internet competition-the model is heavier and enters the upstream of the industry.
In an intelligent cattle farm outside Beijing, the “mystery” of cows’ estrus becomes the code that can be executed accurately.
By combining the camera with AI algorithm, people can monitor the health, physiological cycle, position and other information of the cows, thereby improving the estrus detection rate of the cows, shortening the fetal spacing and improving the reproduction efficiency. At the same time, through the deployment of smart devices such as depth sensors in the cattle farm channel, the health status of the cattle will be fed back to the breeder on time.
Finally, the breeder can achieve “remote cattle breeding” without leaving the house. And these actually start with “pig face recognition”.
Li Jialong recalled the reason for their “pig face recognition”.
At that time, he was responsible for algorithm research at JD.com, in 2017, he received a demand from the insurance department-some farmers will insure a small number of pigs in the family, regardless of which one of the overall died , Will find insurance companies to settle claims, there may be some “deception” behavior.
In the past, the traditional form was to make a mark on poultry, but these marks can be artificially changed. In order to avoid this phenomenon, the company hopes that they will use biological facial recognition technology to accurately identify each pig. That’s how the business around “pig face recognition” started.
The team found that in addition to using AI to serve insurance business, there are more areas in the farm that can be improved. So, in April 2018, Li Jialong and his team were fully stationed in the pig farm, “going and feeding like a breeder.” From office buildings to pig farms, they gradually adapted to the environment mixed with various odors, water vapor, and ammonia with extremely high concentrations. What remains unchanged is to “write code.”
The value of the code was proven in October 2018, “We have received orders for tens of millions of grades for the first time.” “Pig Face Recognition” has also rapidly expanded to areas such as “Bull Face Recognition” and aquaculture.
In addition to JD.com, the story of programmers going to the countryside also happened to technology companies such as Huawei, Tencent, NetEase, Pinduoduo, and Alibaba. Related “farming work” such as nursing for sows, aphrodisiac for cows, and harvesting land are gradually becoming dominated by a group of “polytechnic men”.
Programmer “Going to the Countryside”
Programmers have no wayExpect the farmer master to talk to them about deep learning and wait for others to learn the convolutional neural network and communicate. “The team ’s core personnel, such as R & D personnel, product managers, and designers, go to work with the breeder, what he does, what we do; what he eats, what we eat; and his schedule, we also need to be consistent with him. We treat ourselves as breeders. “Li Jialong and his team used this method to discover problems in unfamiliar fields.
This made them discover something new. “When raising pigs, the breeder usually gets up at 5 o’clock and prepares breakfast for the livestock. It may be busy until 7 or 8 o’clock to feed them their first meal and then go to dinner by themselves. When they come back after dinner, they should feed him The second meal. “This is a very expensive job. “In the past, we needed to mix the feed ourselves, pour it in front of the animals, and feed them. This process was very tiring and troublesome. So, the idea of developing an” automatic feeding system “was derived, which was to do all these things. Automation can be accurately beaten and fed through the control of artificial intelligence.
After the idea was turned into a plan, the breeder reported, “In the past, I had to get up before 6 o’clock, but now it’s just 8 o’clock-just go check and see if the system is working properly.”
The environment of traditional farms is a “nightmare” for newcomers. Due to the presence of various feeds and animal feces, traditional pastures are often filled with various unpleasant smells. The relative humidity of the chicken farm is usually controlled at more than 70%. “The most uncomfortable level is the moment I walk into the chicken farm. The environment is really indescribable.”
Monitoring cow ’s health, physiological cycle and other information through AIoT technology to improve the estrus detection rate of cows , Shorten the fetal spacing, improve reproduction efficiency ｜ Jingdong Mathematics
The problem of poor breeding environment is more severe in the field of aquaculture. “The problems facing aquatic products are not the same as those of cattle, sheep, pigs and chickens. They may face a huge transformation and reform.” Due to environmental issues, some aquaculture models have been banned, such as offshore aquaculture and cage culture in lakes and reservoirs. And freshwater aquaculture.
“In the past, farming in ponds and other environments, the water will come into contact with the soil, some heavy metals and pesticides will remain in the water, and then enter the body of the fish; at the same time, the fish is excreted in the water, and the soil is reversely polluted . “
To resolveIn the above problem, the industry has explored a solution to raise fish in boxes. AI + IoT technology is used to monitor the growth curve, feeding curve and behavior trajectory of fry, as well as various data of water bodies. “In this way, the breeding density can be increased, and it also solves the problems of serious pollution and food safety.”
Traditional aquaculture methods are more violent, and the pollution of water quality and soil is more serious
At this time, the growth process of the animal can be completely controlled by the machine, such as its future weight and nutrients.
In such a business scenario, when a person enters an Internet company, it is no longer an online application scenario represented by APP, but a real physical industry.
Agricultural business of Internet giants
In fact, raising pigs and cattle on the Internet is not a new topic, and the technology giants have already made arrangements. And every family has a regular path to enter the game. Xiao Botian, vice president of North Peak Capital Investment, told geekpark (id: geekpark), “In the agricultural field, technology companies usually output their own technologies and management methods to empower farmers, and thus also achieve their own scale expansion. “.
Similar to JD.com ’s export of smart solutions, Huawei, Alibaba, and Tencent are all taking similar paths. The business of smart agriculture is usually subordinate to the “cloud computing” business department of each company. Almost all of them use cloud computing, big data, artificial intelligence and other technologies to build a “smart agricultural brain”, and then integrate vertical and subdivision directions. The resources of the technology company will eventually form a complete set of integrated solutions and export them to customers.
At present, these customers are mainly government and large agricultural groups. “Our form of cooperation with small farmers is more that the government builds a base. We work with local agricultural enterprises to operate together, and then cooperatives composed of individual households participate in the breeding process. Nature. “Li Jialong told us.
Another typical and well-known “Internet farming giant”, NetEase takes the path of covering the entire agricultural production chain, which is relatively more difficult. E-commerce sales, offline catering channel cooperation, all hands-on. Although from the announcement to enter farming products, to the first own farm put into production, NetEase used for 7 years. But also achieved the feat of 6 people managing 20,000 pigs.
The picture shows a smart farmland monitoring system that can monitor the farm ’s crop growth temperature, humidity, air pressure, light, etc. more than 10 Data for guiding field production.
In the field of planting vegetables and fruits, with big data, artificial intelligence, and the Internet of Things as the technical background, various intelligent solutions have also been born to “control” the growth of agricultural products. For example, the above-ground part is equipped with small weather stations and communication modules; the underground collects light, temperature, salinity and other information through various sensors, and then transmits it to the cloud big data center, and makes decisions and adjustments through the AI artificial intelligence system. Water consumption, fertilizer application.
It can be seen that whether it is vegetables, fruits or pigs, cattle and poultry, agricultural products are moving closer to industrial products, and more of them are born in controllable digital production lines. According to the 2020 Smart Agriculture Strategy Report released by Clear Strategy, the total target market for smart agriculture will reach $ 27 billion by 2020.
Jialong Li judged that there are many companies that are now trying to do the smart agriculture sector. “Everyone has some achievements, but there is no big one. All parties are accumulating their own team and experience resources I personally feel that in less than three years, the field of smart agriculture will usher in an explosion. “
Jiangxi, Jishui County, Jiangxi, the villagers are working in an intelligent seedling raising center, and the working environment is completely different from traditional farmland
However, Xiao Botian believes that the “cake” in this field is not easy to obtain. “Agriculture is an industry with a lot of subdivided fields and subdivided scenes. Smart agriculture can be understood as a collection and its development In a small step and a process of gradual penetration, capital is not the most important link. Data collection, digital technology, professional knowledge in the field of agriculture, data management methods, and industry chain resources are all its thresholds.
In the future, the rise of smart farms is bound to squeeze the living space of small and medium-sized farmers. Xiao Botian said, “In China, the hog breeding industry is gradually shifting from the competitive environment of scattered retail farming to the growing competitive environment of leading companies.”
Logic behind layout agriculture
For technology companies, there are huge business opportunities in smart agriculture, and for companies with e-commerce business, agriculture seems to have become a hot spot. For example, in the past year, fast-handed e-commerce has been signing more original sellers to enter the fast-handed e-commerce ecosystem. At the same time, at the marketing level, the main selling point of “good source”.
In this regard, Xiao Botian believes that the penetration of e-commerce platforms into the upstream of the industry is “a natural thing.” “The e-commerce platform participates in the industry chain through its own channel advantages and appeals, and farmers or agriculture. The Group is working together to create some key categories of products. In this way, by strengthening the upstream control, the quality and price of agricultural products provided to consumers can be more actively controlled, thereby increasing the appeal of the platform to consumers. “
“Good source”, “Straight from the origin”, it is not difficult to see from the promotion level The importance and penetration of the upstream of the agricultural industry
“At a time when competition is becoming more and more homogeneous, in the future, better and more unique products will be more likely to differentiate, which will increase consumer stickiness. This is also a dimension of future competition between e-commerce platforms.” He added.
In addition, the popularization of e-commerce live broadcasts has reduced the cost of farmers ’entry into the e-commerce ecosystem. Farmers can directly connect to the target customer base and eliminate the need for” intermediaries. ” The part that was once “rejected” by the Internet for supply-side merchants due to problems such as operating thresholds also received admission tickets.
This part of the “newly added” businesses are also reflected in e-commerce data. During the Double Eleven in 2019, the sales of Ali’s all-platform agricultural products exceeded 7 billion yuan, an increase of 53% over the same period in 2018. In 2019, the annual turnover of agricultural products reached 136.4 billion yuan, an increase of 109% over the same period of the previous year. The turnover in the first three months of 2020 is close to nearly 70% of the full year in 2019.
A duck farm in Yongzhou, Hunan, employees recommend duck eggs to customers through live broadcasts. Live broadcast e-commerce is becoming important in agriculture Sales channel
The “marginal species” in e-commerce has grown into a mainstream category by leaps and bounds, and there is still huge room for growth.
It is foreseeable that, whether it is to provide technical solutions, sales channels, or jointly create agricultural product brands, the e-commerce platform will be more intensive and fierce in the layout and competition of upstream resources in the agricultural industry.
In this process, farmers and Internet companies will usher in more “intimate” contact. When the boundary between the Internet and traditional industries becomes blurred, the new agricultural production and sales process requires “new farmers.”