In golden September, as rice waves roll and fruits fill the air with fragrance, another Chinese Farmers' Harvest Festival arrives. This year, Jiangsu Province's No. 1 document explicitly proposed promoting "AI plus agriculture" development. The provincial government issued policies for strengthening agriculture through science and technology, and the province deeply implemented the "AI plus agriculture 141" initiative—one scheduling system, four industrial models, and one equipment support system have taken root. Artificial intelligence is rapidly penetrating crop farming, aquaculture, horticulture, and agricultural product processing chains. From the vast fields of Yancheng to the crab ponds of Kunshan, from the smart greenhouses of Haimen to the intelligent workshops of Jiangyan, algorithms have become farming strategies and data has become agricultural capital. These smart production scenarios vividly illustrate the new harvest scene of Jiangsu's modern agriculture.
AI Smart Farming: From Experience-Based to Data-Driven Cultivation
On September 22, in the Lixiahe Plain hinterland, rice at the Qixing Modern Farm in Yandu District, Yancheng City, entered its grain-filling stage, with 5,000 acres of continuous paddies rolling in green waves. A gentle breeze rustled the gradually plumping rice ears, and the air was filled with the fragrance of fresh rice. On the blue screen of the farm's smart management center, satellite remote sensing imagery, drone field patrol footage, soil moisture curves, and pest monitoring data updated in real time simultaneously. Staff member Gu Yingying sat at her computer, clicked the mouse, and pulled up that day's field "diagnosis report." Gu explained that once the drone completes a flight, AI produces a fertilization plan within minutes. This is the typical application of the Qixing Farm's "sky-air-ground-human-machine" five-in-one AI smart farmland system. Since the start of this year, the district has strengthened cooperation with the Provincial Internet Agriculture Development Center and Nanjing Agricultural University, using multispectral drone remote sensing and intelligent nest facilities as the core. By integrating satellite remote sensing, field IoT sensors, and other multi-source information, they developed five AI algorithm models—seedling identification, growth monitoring, variable-rate fertilization, yield estimation, and lodging warning—connecting the entire chain from equipment monitoring to intelligent analysis to prescription generation to plant protection drone execution. This is quietly transforming how this land is cultivated. The change at Qixing Farm stems from a shift in decision-making methods. The entire rice production cycle is now scientifically divided into key stages: sowing, tillering, irrigation, fertilization, plant protection, and harvesting. For each stage, AI models provide prescriptions based on real-time data—what to do and how much to do. "Our five independently developed core AI algorithm models work in coordination, which is equivalent to equipping each field with an indefatigable, precise, and professional 'AI housekeeper,'" said farm technician Ji Xianglong, using an analogy. "It accurately grasps crop growth rhythms, knows when crops need water, when they need fertilizer, and when pests may occur. Relying on real-time data analysis, it provides tailored farming plans, truly achieving one prescription per field, precision to the mu, and measures applied as needed." The data proves the system's effectiveness: it collects field data around the clock, and inspection efficiency is more than 30 times that of manual patrols. Chemical fertilizer and pesticide use has been reduced by more than 15%, while water resource utilization has significantly improved. Average agricultural materials cost per mu has dropped by over 5%, and rice and wheat yields have increased by 8% to 12%. The project was built and put into operation in 2025, creating a 10,000-mu smart demonstration area in Yandu, with the technology model extended to Huai'an, Funing, and other locations. "Currently, the intelligent monitoring and precision operation linkage system covering the entire life cycle of rice and wheat at Qixing Farm has formed a mature solution and has been included in the '2025 National Smart Agriculture Typical Cases Compilation' released by the Information Center of the Ministry of Agriculture and Rural Affairs," said Xie Zhaobiao, director of the Yandu District Agricultural Technology Extension Center. He stated that Yandu will continue to deepen the integrated application of AI in agricultural condition monitoring, pest control, and precision operations, accelerating the creation of more replicable and scalable smart farm models so that data truly becomes the new "agricultural capital" of Yandu agriculture.
AI Smart Aquaculture: AI on Duty in Crab Ponds
As autumn winds rise over Yangcheng Lake, hairy crabs complete their final molting and are storing fat and developing roe, quietly awaiting the harvest that begins in a few days. Entering the Bacheng Town Yangcheng Lake Hairy Crab Industrial Park, water quality sensors and high-definition cameras are scattered along the pond embankments. This is Suzhou's first "5G plus smart fishery" and a provincial fishery smart park in Jiangsu. Core indicators such as dissolved oxygen, water temperature, pH, ammonia nitrogen, and nitrite are transmitted in real time to the big data cloud platform, with all ponds visible on a single screen. Yangcheng Lake hairy crab is a signature industry of Kunshan, with a total farming area of 32,250 mu, an annual output of approximately 15 million premium crabs, and an industry chain annual output value approaching 6 billion yuan. More than 30,000 people earn their livelihood from this "crab rice bowl," and Bacheng Town is known as the "Hometown of Chinese Yangcheng Lake Hairy Crabs." As a national advanced county for agricultural science and technology modernization, Kunshan is gradually "planting" IoT sensing and AI algorithms into crab ponds. "In the past, during typhoons and heavy rain, I feared oxygen depletion and pond overturning in the middle of the night—I couldn't sleep soundly," said Xu Yufeng, who has 26 years of crab farming experience. She uses a solar-powered IoT water quality monitoring system in her 41-mu crab pond. "When dissolved oxygen falls below 5 milligrams per liter, the aerator automatically starts, and if there's a power outage, an alarm is pushed remotely." Wang Changbo, deputy director of the Kunshan Aquatic Product Technology Extension Station, explained that Kunshan is building a smart hairy crab farming platform expected to be fully completed and operational in the first half of next year. It focuses on exploring the industry's missing "brain" module: building crab growth prediction models and video monitoring algorithm models to forecast growth trends combined with time, environment, and feeding conditions; using underwater real-time video to calibrate individual specifications and output feeding strategies; analyzing water quality, aquatic plant species, and distribution to develop improvement and treatment strategies; and linking automated hardware for scheduled, quantifiable automatic feeding. As a Suzhou smart agriculture demonstration base, Jiangsu Jingshang Mantang Agricultural Development Co., Ltd. has not only deployed IoT water quality sensing equipment at its own base but also exported standardized digital management solutions to partner farmers: uniformly collecting production records of pond water quality, feeding, and molting, forming digital archives for each pond, guiding crab seed stocking and feed feeding based on historical data, and incorporating scattered pond farming processes into standardized management. In a few days, the intelligent sorting line in the workshop will begin full-load operation to welcome the harvest wave of hairy crabs. The confidence in the harvest lies in the data transmitted by sensors, in continuously iterating algorithms, and in the hard work and sweat of every crab farmer.
AI Smart Horticulture: A Four-Season Harvest Symphony in Greenhouses
Sensors capture subtle changes in temperature, light, water, and fertilizer in real time. Mechanical arms tirelessly complete sowing and transplanting. Agricultural robots shuttle between greenhouses for iterative trials. On the coast of the river and sea in Haimen, Nantong, traditional farming that once depended on weather is now being reshaped by digital technology, painting a distinctive harvest scene in the land of fish and rice. Entering the A+ Greenhouse Works in Changle Town, the 89-mu park bursts with vitality. This modern agricultural park, with a total investment of 63 million yuan, breaks geographical constraints by bringing southern produce into northern greenhouses, maintaining indoor temperatures at 25 degrees Celsius. In the strawberry cultivation area, newly transplanted strawberry seedlings are neatly arranged. Coconut coir soilless cultivation is used, with drip irrigation pipelines delivering water and fertilizer precisely to each strawberry root. Reflective horticultural fabric on the ground reflects sunlight fully to the lower parts of the plants, promoting healthy growth. "We achieve green planting through ecological methods," said Li Zhijun, the park's operations director, pointing to yellow and blue sticky insect boards hanging in the greenhouse. The park releases predatory mites that feed on spider mites, the natural enemy of strawberries. Varieties such as Hongyan, Tianye, and Heimolin, adapted to local tastes, have made their home here, with the first batch of fresh fruit available for picking in November and peak market supply expected around the Spring Festival. During harvest season, parents bring children with small baskets for picking, creating a rural leisure attraction. Crop after crop, strawberries are followed by tomatoes and cucumbers in rotation, and summer brings hanging melons in succession—inside the greenhouse, harvests continue through all four seasons. The leafy vegetable factory in Zhengyu Town is equally impressive. In the constant-temperature seedling room, seeding machines precisely place seeds into seedling trays. After a month of cultivation, seedlings are transferred by automated equipment to hydroponic floating boards, where nutrient solutions are automatically mixed and supplied according to crop needs, yielding 12 harvests of lettuce per year. "Normally only one person is needed for operation and maintenance, with three additional workers during harvest and packaging," said Li Suting, a staff member of the Zhengyu Town Agricultural Development Office. The fully enclosed production environment reduces external pest invasion and lowers pesticide application frequency. When individual lettuce heads reach the order standard of 130 grams, they are immediately harvested and packaged, reaching the table within two hours from the greenhouse. The park has now established a digital agriculture platform with real-time environmental data monitoring, and the next step is upgrading equipment ports toward full-process remote control. While the planting side continues to iterate, local Haimen enterprises are tackling agricultural robots, pioneering "machine substitution" for facility agriculture. At Jiangsu Zhenghai Intelligent Equipment Co., Ltd.'s production workshop, engineers are busy debugging multiple picking, inspection, and targeted spraying robots. These robots can identify the maturity status and pest damage points of fruits and vegetables such as tomatoes, strawberries, and hanging melons, while protecting fruit integrity to the greatest extent during picking. Not relying on weather, but creating four-season produce through wisdom. In Haimen, smart greenhouses, plant factories, and agricultural robots complement each other, redefining the four-season farming approach. Human-machine collaboration and technology-assisted agriculture are writing a new chapter of harvest stories for the new era through repeated trials and refinements.
AI Smart Processing: Agricultural Products Turned into Delicious Beverages with One Click
What complex processes are required to turn agricultural products grown in the fields into delicious beverages? Jiangsu Shangshou Biotechnology Co., Ltd.'s answer is "one click." As a provincial-level leading agricultural enterprise, Shangshou Biotech focuses on "AI plus agricultural deep processing," systematically integrating AI technology into production scenarios to achieve intelligentization across the entire chain, from raw agricultural material handling to finished product warehousing. Walking into the Shangshou Biotech facility in Dalun Town, Jiangyan District, Taizhou City, several large trucks are waiting to load at the entrance. The shipping area is neatly stacked with cases of beverages—coconut water, red bean drinks, fruit juices, flower teas, and many more, totaling over a hundred varieties. "Our shipment volume is large, running at full capacity year-round, with intelligent production lines operating 24/7 without stopping," said Miao Jianming, the company's deputy general manager. Inside the production workshop, several intelligent lines are running at full speed to fulfill orders. Viewed from above, pipelines run up and down, connecting various equipment and operating systems, with several tall tanks standing in several locations. The entire production line is precisely arranged, resembling an intelligent workshop with high technological content in the industrial sector. At the back end of the line is the beverage conveying section, where bottles of "jasmine ice coconut water" queue neatly along conveyors toward the packaging area. The long conveyor belts weave through multiple curves, yet the transport speed remains swift. Only a few workers appear in this area of the production workshop. "They are mainly responsible for promptly collecting and handling 'defective products' that fall off the line," explained Qi Chengqi, deputy workshop director. AI algorithms and recognition systems are also introduced here. Beverages passing through the conveyor undergo multiple inspections—products deemed "unqualified" by the system, such as improperly labeled or imperfect bottles, are decisively "kicked" off the conveyor by AI. Currently, more than 160 mu of premium red beans in Shenyang Village and Xingchi Village of Dalun Town have entered the harvest period, and their next stop is Shangshou Biotech for processing into beverages. "Through village-enterprise joint construction, we mobilized villagers from both villages to plant red bean varieties designated by Shangshou Biotech, supplying them to the enterprise on an order basis," said Gu Mingzhu, deputy mayor of Dalun Town. The enterprise's purchase price is slightly higher than market rates, increasing farmer income by tens of thousands of yuan annually. In Jiangyan, "AI plus agriculture" is transforming toward full-chain penetration. Jiang Wei, Party secretary of the Jiangyan District Agriculture and Rural Bureau, introduced that the district will rely on the construction of regional comprehensive agricultural service centers, continue promoting the deep integration of AI from agricultural production to processing stages, cultivate smart scenarios according to local conditions, and consolidate the harvest foundation through science and technology.
When Harvest Meets Algorithm
On September 21, farmers were picking kiwifruit at a planting base in Jijiazhuang Village, Shizong Town, Tongzhou District, Nantong City. During the golden autumn season, Jiangsu presents a different harvest scene: algorithms are becoming a new type of "agricultural capital" empowering modern agriculture. Looking back at the long river of farming civilization, "depending on the weather" has been written throughout agricultural history. The 24 solar terms are our ancestors' simple decoding of natural rhythms, and agricultural proverbs are the accumulated experience passed down through generations. But even the most seasoned farmer cannot break through the limitations of individual experience. When will pests break out? What is the optimal harvest date? How much fertilizer should be applied? These questions that have troubled farmers for thousands of years often have answers that are merely "about right." And the accumulation of "about right" becomes the ceiling on yields and a bottomless pit of costs. We have accumulated more varieties, better machinery, and more efficient pesticides, but the core decision-making mechanism still relies on human observation, judgment, and intuition—until algorithms entered the picture. The essence of algorithms is converting "experience" into "data" and structuring "intuition" into a reproducible decision-making logic. The experience an agricultural technician accumulates over a lifetime might be traversed, learned, and surpassed by AI within seconds, because what AI consumes is not just one person's experience but big data from countless fields, years, and variable combinations. At the Qixing Modern Farm in Yandu, Yancheng, AI intelligent prescription maps guide variable-rate fertilization with precision. In Kunshan's ecological crab ponds, intelligent sensing equipment linked to aeration systems operates 24/7 to safeguard farming safety. In Nantong Haimen's leafy vegetable factory, algorithm-controlled light, temperature, water, and fertilizer achieve 12 standardized harvests per year. In Taizhou Jiangyan's intelligent processing workshop, AI visual recognition systems accurately identify defective products. Data is no longer abstract parameters detached from the fields, and algorithms are no longer virtual models in laboratories—they are deeply integrated into the entire chain of plowing, sowing, managing, harvesting, and processing as new production factors. For millennia of farming civilization, farming decisions depended on farmer experience, seasonal patterns, and field observation. Jiangsu's smart agriculture practices, relying on multi-source data including satellite remote sensing, IoT monitoring, and drone surveys, deeply analyze crop growth patterns and field environmental changes through AI models, customizing exclusive management plans for each field, advancing from "farming by experience" to a new stage of "deciding by data and improving quality through algorithms." The "AI plus agriculture 141" initiative being promoted across the province holds its value precisely in building a new agricultural knowledge production and distribution system—gathering the experience scattered in agricultural technicians' minds, the mechanisms written in papers, and the patterns precipitated in data into algorithm models, then "distributing" them through phone screens to every farmer. In the past, farmers in remote villages might never encounter top agricultural experts in their lifetime. Now, opening a mini-program provides decision-making support equal to that of demonstration households. However, algorithms also have limitations. The agricultural data environment is far more complex than industry—the climate changes, soils vary, and pests differ year by year. A model that performs excellently in southern Jiangsu may need retraining in northern Jiangsu. High-quality, fully annotated, and full-scenario agricultural datasets remain scarce. Facing reality, Jiangsu's "AI plus agriculture" is still in the stage of deep cultivation and iteration. Intelligent hardware equipment is becoming increasingly mature, but core algorithms adaptable to complex field scenarios with autonomous judgment and decision-making capabilities are still being refined. Precisely because it is known to be difficult, the exploration is all the more valuable; precisely because the journey continues, a coordinate is needed to mark the direction. The Harvest Festival, now in its ninth year, can serve as this coordinate, marking the transformation scale of Chinese agriculture from existing to excellent, from excellent to strong. This year's marker, Jiangsu has written the keyword "AI": replacing manpower density with data density, compensating for resource insufficiency with algorithm precision, and breaking the path for high-quality development of modern agriculture. When harvest encounters "algorithms," we usher in a new era where experience and data dance together, humans and machines collaborate, and tradition and future converse. The warmth of algorithms resides on the ridges, by the crab ponds, and in the relaxed brows of every farmer. That is the most touching encounter between technology and the earth, and it is also the most solid answer Jiangsu has written for this harvest festival.
Comments