Regaining the flavor of childhood tomatoes | Image source: pixabay.com
Written by | Li Cheng
● ● ●
"The world's largest vegetable crop" is actually tomatoes?
This small wild fruit, which was taken out of the Andes by the Spanish colonists in the early 16th century and neglected, has been planted in the world for about 5 million hectares and has an annual output value of about 100 billion US dollars, with tens of thousands of varieties, as well as low calorie, rich in antioxidant lycopene and a variety of minerals, vitamins and other advantages, conquered consumers around the world, becoming the "chosen son" of vegetable crops.
China is the world's largest tomato planting area (planting area of 16 million mu), the largest yield of the country, the annual output of about 55 million tons, accounting for about 7% of the country's total vegetable production. However, how to make tomatoes more high-quality, higher yields and more nutritional value is still the goal pursued by the world's top agricultural researchers.
At the second "Duoduo Agricultural Research And Technology Competition" that just concluded on April 25, 2022, people were pleasantly surprised to find that a multinational youth team of "Tomato Fast Growing" won the championship with a yield and quality that far exceeded that of traditional cultivation. Using the "human-machine fusion intelligence" algorithm, the growth of tomatoes has become a vivid strategy game, and the yield per unit area of cherry tomatoes at the end of the event is 11.3 kg/m2, which is expected to be close to 30 kg/m2 at the end of the normal growth cycle, which is comparable to the peak yield of small tomatoes grown in the world's leading Dutch greenhouses at the same time.
The competition was jointly sponsored by Pinduoduo, China Agricultural University and Zhejiang University, and received technical guidance from the Food and Agriculture Organization of the United Nations and Wageningen University and Research Center in the Netherlands, and the world's top agricultural research teams used interdisciplinary planting technology and AI algorithms in the competition to grow crops with high quality and high yield in accordance with sustainable planting practices, so that people are full of endless imagination for the future of smart agriculture.
1
Dialogue with plants: human-machine fusion of intelligent vegetables
Figure 1 The "Duoduo Agricultural Research and Technology Competition" competition base in Kunming, Yunnan Province, | Photo: Peng Lai
At an altitude of about 2,000 meters above sea level in Kunming, Yunnan Province, four glass greenhouses with an area of 96 square meters are lined up, and the stage of the agricultural research and technology competition for tomato cultivation is located here.
The competition provides a separate greenhouse and corresponding environmental control application interface for the finalists. Since October last year, the four teams have been planting tomatoes in Beijing, Shanghai, Zhejiang, the Netherlands and other places through software remote control of the intelligent greenhouse located at the Kunming competition base, combined with precision agriculture technologies such as nutrition, computer vision and deep learning.
During the 186-day competition, whether it is the "Digital Tomato Intelligent Control Cloud Platform", "CO2 Enrichment Technology", or the "Disease Management System", as well as the automatic identification system for the growth status of tomatoes, people have sighed one after another: our cognition of smart agriculture is somewhat "unknown to Han, no matter Wei Jin".
The greenhouses in Kunming, Yunnan are mostly single plastic greenhouses, and the traditional planting methods are backward in planting technology and have a large demand for labor. In this competition, a wisdom greenhouse was built, the basic hardware generally includes planting troughs, coconut bran strips, rock wool matrix blocks, etc., different from ordinary soil, tomato planting matrix physical and chemical properties are stable, not easy to decompose, rot or denatured, with good water retention and breathability, low cost, light quality, but also support controllable adjustment, is the best partner for precision planting. In addition, because the solar greenhouse or plastic greenhouse used in the covering facilities will greatly reduce the ultraviolet rays reaching the inside of the greenhouse, affecting the growth and development of vegetables, while the smart greenhouse regulates the indoor temperature and ultraviolet rays through the heating system and LED fill light, so as to promote plant growth and fruit growth and development.
Outside of hardware, the core of technology is software: environmental control applications. In order to plant a cherry tomato, the contestants used sensors, spectra, algorithm models, and even remotely adjusted the greenhouse water fertilizer, light, carbon dioxide and other growth environment elements, "people sit at home, tomato shed long", to achieve no one planting, full of futuristic sense.
In the past, cherry tomato growth status identification and identification relied entirely on the experience of growers, and the cherry tomato growth status recognition system based on machine vision can effectively help solve such problems.
The "Zhijia" team led by China Agricultural University created a "digital tomato intelligent control cloud platform" that can "dialogue" with plants, which established a tomato plant growth environment model including net photosynthetic rate model and transpiration rate model, and introduced the relationship parameters of environmental and plant information interaction. When the real-time monitoring data of the environment exceeds the threshold range, the system platform will send SMS reminders to the operators to form a closed-loop intelligent control.
Tomatoes require more water, and irrigation strategies vary according to growth periods and growth states. The "Zhijia" team uses machine vision and sensor technology to identify the physiological information such as leaves, flowers and fruits of cherry tomatoes in real time, and conducts real-time monitoring of ambient light conditions and temperature and humidity, and then calculates the transpiration rate of tomatoes through the leaf area of tomatoes and real-time photosynthetic effective radiation, and then derives the water demand of tomatoes, so as to remotely adjust the amount of water and fertilizer irrigation and irrigation frequency.
Figure 2 Digital Tomato Intelligent Control Cloud Platform developed by the "Zhijia" team
Led by Beijing Polar Star Agriculture and composed of researchers from China Agricultural University, Northwest A&F University and Wageningen University in the Netherlands, the "Tomato Fast Growing" team adopts a ring control system that includes temperature, humidity, light, ventilation, curtain and carbon dioxide and all the environmental parameters required during tomato growth. In terms of irrigation, they mainly adjust based on the amount of light accumulated: when the amount of light accumulates to a certain value, an irrigation group is started for irrigation. But at the same time, they also installed their own matrix humidity and EC sensors to monitor root moisture content, and once the matrix moisture content is found to be too low but has not yet triggered irrigation based on light accumulation, they issue a supplementary irrigation instruction through the water and fertilizer system.
Seemingly simple window ventilation, in fact, need to consider many factors at the same time, including time, wind direction / wind speed, whether it rains, whether extreme weather, etc., but also consider the internal and external temperature difference, light, humidity and other effects on the window speed, size and direction. It's hard to imagine how a common ordinary greenhouse can cope with so many factors.
The "Tomato Fast Grow" team has been able to keep the tomatoes at the right growing temperature through a variety of ways, such as window opening, fan circulation, and high-pressure spraying devices. The team even set up different window opening strategies for the headwind side and the downwind side, such as the window opening strategy of the headwind side is more conservative, in order to prevent a large amount of cold air from entering the greenhouse instantaneously, causing a sudden drop in temperature and triggering a serious plant physiological reaction.
In addition, appropriately increasing the relative humidity of the growing environment can significantly increase the tomato leaf area, however, too high humidity can also lead to a series of problems, condensate will spread viral and bacterial diseases, high humidity will trigger the germination conditions of fungal diseases of spores. So the team triggered the opening and closing of the sprayer or window in real time for different humidity and different temperatures, down to every degree and day and night. And all of this is basically remotely controlled.
Such precise growth control completely bids farewell to the dilemma of traditional agriculture "relying on the sky to eat".
2
Smart Greenhouses focuses on sustainable development
Figure 3 The finalists of the 2nd "Duoduo Agricultural Research and Technology Competition" | Photo: Peng Lai
"Silent Spring", published in the 1960s, sounded the alarm for mankind, and excessive reliance on chemical pesticides to control crop diseases and insect pests will cause a large number of problems that are difficult to repair, such as fruit pesticide residues, increased resistance of pathogenic bacteria and environmental pollution. The advantage of the smart greenhouse is not only in intelligence, but also pays more attention to environmental friendliness and human friendliness, and can solve the problem of pesticide residues in vegetables and fruits to the greatest extent from the source.
The four teams that reached the finals all offered different ideas for pest control. The "Xi Persimmon" team composed of young scholars from Zhejiang University believes that in actual agricultural planting, accurate identification of disease types is an important link, and they have designed and implemented tomato disease classification and disease positioning algorithms, which have achieved an accuracy rate of 98% in the identification of eight diseases such as early blight and powdery mildew, and then help provide corresponding prevention and control strategies.
The "Nongsheng Brain" team from Jiangsu University, Beijing Municipal Agricultural Technology Extension Station, Beijing Technology and Business University, Jeonju University and other scientific research institutions and colleges, with diversified scientific research backgrounds, eliminates the occurrence of soil-borne diseases such as tomato root rot, blight and anthrax from the source by selecting high-quality healthy seedlings. In the tomato production process, we should focus on the prevention and control of airborne diseases, matrix-borne diseases and insect pests, and control the diseases and insect pests at a very low level before flowering, so as to reduce the occurrence of diseases and insect pests during flowering, and avoid using drugs during flowering to avoid hurting flowers. In the presence of pests and diseases, yellow sticky worm plates are used to trap aphids and whiteflies, and blue armyworm plates are used to trap thrips, and the green is pollution-free and the cost is lower.
And tomato planting common fungal diseases such as powdery mildew, leaf mold, black leaf mold, gray mold, etc., its spore infection period, incubation period human naked eye can not identify, and so crops have obvious symptoms of infection, the fungus has been in the outbreak period, produced a very large number of new spores, this time to intervene, the loss is larger, the cost is also higher.
The "Tomato Fast Growth" team predicts the risk of contracting common fungal diseases with 89% accuracy by developing a disease management system composed of multiple machine learning models. In the stage of spore infection of crops, the disease management system can predict the risk of disease occurrence and intervene in advance, such as taking precise preventive spraying drugs, changing environmental control measures and other means to minimize the risk of disease occurrence, which can effectively reduce pesticide consumption and protect the environment. Ultimately, the tomatoes produced by the team also did not detect any pesticide residues.
The "Zhijia" team uses the identification method of machine vision to automatically identify tomato diseases, and the results of the identification and analysis are transmitted to the disease expert system for decision-making, and if it is indeed a diseased leaf, an alarm message will be issued and the corresponding expert solution will be given.
During the planting process, the teams also fully considered other aspects of cost and environmental issues, using various refined means to save energy. For example, the "Xi Persimmon" team used a modified zeolite molecular sieve, which has stronger adsorption capacity and slow dissociation capacity for CO2 in the lower concentration range, which promotes the improvement of photosynthesis while effectively avoiding the secondary emission problem caused by direct co2 application and limited plant utilization. They have also developed a light simplified CO2 enrichment device for glass greenhouses and plastic greenhouses based on zeolite molecular sieves, which is simple and efficient, and the materials can be reused and reused, which is a new direction to promote the development of low-carbon agriculture.
Proper supplementation of light when the light is insufficient can significantly increase the yield of tomatoes. However, considering the high energy consumption of the fill light lamp, the "Tomato Fast Growth" team adopted a double-layer light filling strategy, and the crop model judged the photosynthetic efficiency of the plant according to carbon dioxide, external light and temperature conditions, so as to determine whether to turn on the light. Therefore, the production cost of the "tomato fast growth" team in winter is only 15.3 yuan / kg, and according to industry experience estimates, the average production cost of small tomatoes harvested in the same season is about 20 yuan per kilogram.
In order to maximize the economic operation benefits on the basis of achieving the expected planting effect, the "Zhijia" team comprehensively analyzed the tomato plant growth model, environmental prediction model and cost model, and adjusted the greenhouse environmental parameters and operating costs in combination with the calculation results. The "Xi Persimmon" team strives to be the first in this regard, formulates a precise resource consumption management strategy, and realizes on-demand supply by building an accurate tomato demand model, and is in a leading position in terms of water consumption, fertilizer consumption, electricity consumption, labor, plant protection, seedlings and other material resource consumption, and performs best in energy consumption control.
3
Regain the tomato flavor of childhood
How many people can't forget the "persimmons" of childhood, break open, powder the sand, take a bite, the rich tomato flavor straight to the brain, if you cut the pieces and mix with white sugar to eat, it is to lick the soup left at the bottom of the plate clean. Today's tomatoes, all red and round and rolling hard, beautiful, but the taste is dull. People's impression of modern agriculture is often that the production capacity is getting stronger and stronger, the output is getting higher and higher, but the flavor is sacrificed, and the taste of childhood can never be found again.
As an important facility vegetable crop in the world, tomato production is becoming saturated year by year, and people are increasingly pursuing its taste quality and nutritional value, and the direction of modern agricultural efforts is also changing.
Fig. 4 Cherry tomato | Photo by Markus Spiske on Unsplash
The taste of tomatoes is intuitively reflected by the sugar content and sugar-acid ratio, which is the ratio of soluble sugar and organic acid (malic acid/citric acid). The higher the sugar-acid ratio, the sweeter the tomato feels to eat, and vice versa, the more sour it is.
Well-known cherry tomato varieties on the market, such as "zhi pure" soluble sugar content of 8.5%, "summer sunshine" is 9.3%, after the precise control of tomato growth, the "tomato fast growth" team of cherry tomato average soluble sugar content of 9.75%, and the sugar fluctuation range is small throughout the planting period, the quality is stable. Their cherry tomato has a comprehensive sugar-acid ratio of 13.3, which is beyond the sugar-acid ratio of ordinary tomatoes (6-10) and even close to the sugar-acid ratio of red strawberries (13.9). Vitamin C, lycopene, amino acids, etc. also have a very good content, of which the content of vitamin C is 35.1 mg /100 grams, far exceeding the vitamin C content of ordinary tomatoes.
The "Xi Persimmon" team, which is better in flavor and nutrition, found that the morning light supplementation can increase the content of soluble substances such as glucose, sucrose and fructose by increasing enzyme activity, so in order to cope with the problem of low light intensity and short time in winter, the team combined with the algorithm to develop a fine light control program. At the same time, on the basis of the preliminary research on the influence of different water and fertilizer formulas on the comprehensive nutritional quality of cherry tomatoes, combined with the results of the water sample analysis at the competition site, the water and fertilizer formulas required to achieve the best nutritional quality of tomatoes in various periods in the field environment were determined, the irrigation model of "three-stage" management was adopted, and the latest research results such as plant sulfonopeptide (PSK), fish protein, trehalose, and plant prebiotics (rapelide, BR) were applied to maximize the nutritional quality and taste value of tomato fruits. The average soluble solids of the entire growing cycle of the cherry tomatoes they grew averaged 9.8 percent, an 80.8 percent increase in vitamins, 72.5 percent in amino acids, and 114 percent more in lycopene than commercially available traditionally cultivated tomatoes.
4
The vast world of new farmers after the 90s
The 4 teams shortlisted for the "Duoduo Agricultural Research and Technology Competition" accounted for about 80% of the post-90s and post-95 members, and the younger generation of agricultural science and technology workers have become the main force of the competition, which left a deep impression on the judges.
Zhu Chang'an of the "Xi Persimmon" team is a second-year PhD student in the College of Agriculture and Technology of Zhejiang University, engaged in research on the regulation of growth and development of facility horticultural crops. He grew up in the countryside, and he was accustomed to seeing adults using their own hands to make a living in the loess soil, and the high-intensity labor was only a barely enough life, which deeply stung his heart.
At Zhejiang University, Zhu Chang'an met many like-minded friends, some of whom excavated functional genes from the wild tomato genome to improve existing tomato varieties, some of them studied new facilities of horticultural carbon dioxide enrichment equipment, some developed new sensors to detect the stress state of plant bodies, and some applied functional substances to improve the nutritional and flavor quality of tomato fruits. After learning that the second Duoduo Agricultural Research Technology Competition was about to start, this group of young people with their own skills immediately formed a team to participate, and invited some friends in the field of artificial intelligence algorithms and agricultural Internet of Things to join the team.
Figure 5 Xu Dan (second from right), captain of the champion team "Tomato Fast Growth", | with the team members Photo: Peng Lai
Xu Dan, a graduate of Wageningen University in the Netherlands, a top agricultural institution in the Netherlands, has been growing tomatoes for more than two years at one of the largest tomato growing enterprises in the Netherlands. Large greenhouses, orderly production, various environments, automated control systems such as water and fertilizer, and automated logistics sorting, etc., all opened his eyes, and he deeply realized: "The gap between us and developed countries is so big!" ”
In 2015, Xu Dan returned to China and successfully landed the production technology of the Dutch large greenhouse in China, hoping to let more young people join the team of modern agriculture. But the harsh and realistic question is: How to quickly cultivate new growers? How to quickly adapt to the growth needs of different crops in different regions of China? "The Toto Agricultural Research Competition is a bold attempt to join crop modeling experts and data scientists in exploring how to use artificial intelligence to enable growers to efficiently manage a greenhouse, master skills more quickly, and manage more parks across space and time constraints." During the 6-month competition, Xu Dan's "Tomato Fast-Growing" team's human-machine fusion algorithm helped them complete the challenge almost completely remotely.
Figure 6 Human-machine fusion algorithm developed by the "Tomato Fast Growth" team
Led by Zhao Ran, an associate professor at China Agricultural University, the team of 10 people is studying for a Ph.D. or master's degree. From the first sight of the greenhouse to the end of the competition, they constantly changed roles between scientists, workers, programmers, as a "student team", planting experience is less, but through the team's efforts and cooperation, they still achieved a good result of second place in effective yield.
When he tasted the tomato that integrated technology, Zhao Ran felt that everything was so worth it. Zhao Ran also found that the teams entering the finals are very young, which he believes sends a positive signal: more and more young people are willing to engage in agriculture, and China's agricultural modernization is inseparable from young people with solid theoretical knowledge and rich practical experience.
Zhu Qingzhen of the "Nongsheng Brain" team is exactly the type of young man Zhao Ran describes. He was once confused about "what kind of person he wants to become in the future", and after deeply understanding the backward development of agricultural technology and equipment in the mainland, he resolutely left an automobile design and research and development enterprise and turned to the field of agricultural science and technology. Zhu Qingzhen, who grew up in the countryside, hopes to promote the mechanization and modernization of China's agriculture through his own efforts. After receiving his Ph.D. in 2019, he joined Jiangsu University to engage in teaching and scientific research related to precision agriculture.
"How to empower traditional agricultural industries with cutting-edge technology?" It is a problem that has always haunted Zhu Qingzhen's mind, until I heard that the proposition of the second Duoduo Agricultural Research And Technology Competition is "using diversified artificial intelligence related technologies, challenging the cultivation of high-quality, high-yield tomatoes in a green and sustainable way, and verifying the feasibility of commercialization", just like a beacon in the fog, zhu Qingzhen suddenly lit up, "This is the area we want to explore and deepen, we have set up a nongsheng brain team that includes 3 doctors and 6 masters to register for the competition." Each team member has a wealth of experience in agricultural science and technology. ”
At the same time, Zhu Qingzhen also plans to share the experience of this competition as a political education class to students, attract more young people to enter agriculture, and encourage them to write their scientific research results on the vast agricultural land.
"The introduction of artificial intelligence into agricultural production can significantly reduce production costs, liberate labor, increase production capacity, and improve the stability and uniformity of agricultural product quality." He Yong, a judge of the competition and dean of the School of Biosystem Engineering and Food Science at Zhejiang University, affirmed, "It is worth our joy that Internet technology has found a breakthrough in agriculture. As a bridge connecting agricultural scientists, digital technology experts and farmers, the Duoduo Agricultural Research Technology Competition, a top competition for agricultural researchers, has promoted mutual understanding between technology developers and technology users, and effectively improved the application rate of new technologies. ”
"We are using the technology we have learned to transform agriculture. Big data and artificial intelligence are subverting this traditional industry, the production method of 'face to face loess and back to the sky' will no longer exist, and in the future, only one or two people will be able to manage a vegetable production park." Zhu Chang'an said. After 6 months of training and iteration, their system has achieved very good accuracy and performance.
The champion team of the competition, "Tomato Fast Growth", said in the final summary that they will precipitate the expert cash crop management strategy and AI algorithm applied in the competition into a set of potential solutions - human-machine fusion planting solutions, "We will conduct further feasibility tests in Beijing and several other cooperation bases in the next step, continuous optimization, and hope that in the near future, we can promote the application to the whole country, so that different growers in different regions and different types of facilities can benefit." Captain Xu Dan even thought of a three-stage commercial promotion plan, full of confidence in the future.
Zhao Chunjiang, an expert in agricultural informatization, an academician of the Chinese Academy of Engineering, and also the leader of the judging group of the competition, felt that the young and promising generation was full of enthusiasm and investment in agriculture, and was more confident in the development of China's agricultural science and technology. "The production mode represented by smart agriculture can make production more productive and wiser, let farmers have more income, and let consumers eat more nutritious agricultural products", "Excellent young scientists can put career development in the vast world of agriculture, which will make you make great achievements."
Plate editing | Ginger duck