September 23 this year marked China's ninth "Chinese Farmers' Harvest Festival." During the "15th Five-Year Plan" period, China will strengthen support for agricultural science, technology, and equipment, raising the comprehensive mechanization rate for crop cultivation and harvesting to above 80%. As autumn harvest unfolds, reporters visited fields, warehouses, and workshops to see how various regions are focusing on key scenarios and leveraging intelligent equipment to improve quality and efficiency, transforming traditional autumn harvests into smart bumper harvests.

Zaoyang, Hubei — "Keep straight lines, control spacing, with row spacing errors kept at the centimeter level"

During the autumn harvest season, rice paddies turn golden. In the morning, as dew just evaporated from the rice fields of Qifang Township in Zaoyang City, Hubei Province, Tian Hui drove his harvester into the field. "We can harvest an average of 60 mu per day, harvesting cleanly and more precisely," Tian Hui said.

Tian Hui is a machine operator at the Shunhui Agricultural Machinery Professional Cooperative in Zaoyang City. The secret to his efficient work lies in the cab of that harvester. "This is the BeiDou intelligent terminal for the harvester," Tian Hui said, pointing to a square box installed in front of the control panel. "During operation, by inputting data, it can guide us in real time to keep straight lines and control spacing, with row spacing errors kept at the centimeter level."

The Shunhui Agricultural Machinery Professional Cooperative was founded in 2010. "At the very beginning, we only had 3 machines," recalled Shen Hui, head of the cooperative. Due to low horsepower, limited intelligence, and heavy workloads, the entire team often worked around the clock for over ten days during busy farming seasons. With the implementation of national agricultural machinery purchase subsidy policies, Shen Hui took the lead in upgrading equipment, and members followed suit in acquiring new machines. Today, the cooperative owns 156 various types of agricultural machinery.

Walking into the cooperative's dispatch center, the large screen covering the entire wall displays the trajectories of agricultural machinery at a glance. "All machines are equipped with BeiDou intelligent terminals connected to the smart agricultural machinery management platform," Shen Hui said. "In the past, we only carried out wheat harvesting operations. Now, for crops like rice and rapeseed, we can provide full mechanized services from plowing to planting, and from management to harvesting."

By the end of 2025, Zaoyang City's total agricultural machinery power reached 1.9336 million kilowatts, with over 2,300 BeiDou agricultural machinery operation monitoring terminals installed, and the comprehensive mechanization rate for major crop cultivation and harvesting reaching 92.5%.

Hengyang, Hunan — "Training exclusive drying curve models for over 20 rice varieties"

In Hengyang County, Hunan Province, trucks loaded with rice were queuing to enter the premises of Jiaoshan Rice Industry Co., Ltd. "This is one of the largest raw grain drying centers in Hunan Province, processing 6,000 tons of raw grain per day," said Qiu Suyun, head of the marketing department at Jiaoshan Rice Industry, upon entering the factory area. The entire process—testing, drying, and storage—is connected to the "Big Horned Frog Smart Grain System."

Liao Libao, a major grain grower from the Hengdong Hongyang Planting and Breeding Cooperative, transported freshly harvested rice from the previous night here and handed it over to the production line. The visual recognition system quickly completed a three-dimensional model of the vehicle's grain and then randomly generated sampling points. "Based on appearance alone, the system can intelligently identify over 20 mainstream rice varieties grown by surrounding farmers," said Peng Jiaowen, vice president of Jiaoshan Rice Industry. Cameras conduct 360-degree high-precision scans of the rice from 3 angles, and dozens of features that are difficult for the naked eye to distinguish—such as grain shape, husk texture, and tip shape—serve as the system's benchmarks.

Once testing is complete, the wet grain is transported to the unloading area. "From the moment testing is finished, the destination of the rice is already determined," Peng Jiaowen said. Behind the unloading port, 180 wet grain silos store the collected rice in separate bins according to variety, moisture, and other indicators. Through pipelines, 92 dryers and 132 dry grain silos are connected to them.

"If rice with different moisture levels is mixed together using the same drying temperature curve, it not only dries unevenly but also easily loses weight, freshness, and taste," Peng Jiaowen said. The smart grain system features automatic bin assignment and "one variety, one drying" management. Relying on 24,000 sensors and 260 programmable control modules in the workshop to collect real-time data across the entire chain, it generates optimal drying plans.

"Since the gradual application of the smart grain system in 2023, the system has iterated year after year and has now entered version 3.0," Peng Jiaowen said. "On the basis of training exclusive drying curve models for over 20 rice varieties, the system intelligently adjusts parameters to achieve variable-temperature drying, increasing the rice yield rate by 1% and the head rice rate by over 3% compared to last year's previous-generation system."

Xiaoshan, Zhejiang — "From transport to storage, AI is used throughout"

Xu Jianxiang, a major grain grower living in Yinong Township, Xiaoshan District, Hangzhou City, Zhejiang Province, planted 340 mu of late rice. In previous harvest seasons, he was busy running around, but this year has been much easier. "From transport to storage, AI (artificial intelligence) is used throughout—it's very convenient," Xu Jianxiang said. He took out his phone, opened the "Xiaoliang Purchase and Sales" official account, and after making an appointment, he only needed to deliver the wet grain to the drying service point; the subsequent process required no worry.

"In the past, after grain was dried at the service station, it still had to be sent to grain depot sites to queue for purchase," said Zhao Binjie, deputy general manager of Xiaoshan Grain Purchase and Sales Company. The "Xiaoliang One-Stop Service" AI system can plan the optimal locations online for farmers based on real-time data. Each grain depot can also rely on algorithms to automatically complete vehicle diversion, berth pre-allocation, and staggered equipment scheduling, increasing the average transport efficiency of grain depots in Xiaoshan District by 30%.

In the sunlight quality inspection room, dried rice samples have their own "ID cards." After inspectors read information via barcode scanners, a dedicated file is immediately generated in the sunlight quality inspection information system of the "Zhejiang Grain Depot" digital collaborative application platform for food security. The tall flat warehouse grain depot can hold 6,925 tons of rice and is equipped with AI technology internally. Through technical means such as image recognition and IoT direct grain temperature collection, key values including humidity, temperature, and nitrogen content inside the depot can all be uploaded to the cloud in real time.

During the critical collection and storage period, drones equipped with various sensors undertake multiple tasks including depot area patrols, warehouse roof inspections, and auxiliary detection of gas concentrations inside warehouses. Data is transmitted back to the "Zhejiang Grain Depot" platform, and through AI analysis, staff can immediately locate suspicious marks in the footage and then go to the site for verification and handling, forming a three-dimensional inspection model of human-machine complementarity and achieving 24-hour real-time monitoring of dynamic changes in grain and depots.

Currently, the "Zhejiang Grain Depot" platform's business has covered 232 collection and storage enterprises across the province, with networked access to over 590 depot sites and 6,750 warehouses, connecting the data links for grain collection, storage, transportation, and management.

Hunchun, Jilin — "Judging quality now has a 'sharp eye'"

By the Tumen River, golden waves of rice roll layer by layer. These days, Cui Yuanhao has been busy nonstop, inspecting fields and checking the factory, doing sufficient preparation to welcome the rice harvest. "We hope for a bumper yield, and even more for increased income," said Cui Yuanhao, head of Hunchun Forest Mountain Agriculture Co., Ltd. in Jilin Province. He is very confident in the factory's automated processing production line. "From husking, brown rice separation, color sorting, polishing to vacuum packaging, rice is entirely completed by mechanical automation, and now it is moving toward intelligence."

Stopping in front of a medium-sized machine, the roaring mechanical sound was accompanied by gusts of wind. Husked rice rolled down from the upper part inside the machine, and rice with normal color entered the next process, while a small number of darker grains were screened and blown out by the machine. "This is an intelligent color sorter. Photoelectric equipment precisely identifies rice color, giving us a 'sharp eye' for judging quality," Cui Yuanhao introduced.

The processing plant covers nearly 10,000 square meters and has a daily processing capacity of 150 tons, five times that before renovation and upgrading. During processing, generally only 3 people are needed: one to watch the raw materials, one to manage the machines, and one to monitor the finished products. The upgraded production line can not only produce traditional white rice but also products such as germ rice and brown rice.

Jilin Province has continued to promote smart agriculture in recent years. With policy support, in 2023 the enterprise where Cui Yuanhao works underwent comprehensive upgrading and renovation, incorporating automated and intelligent design, driving simultaneous improvement in benefits. In 2025, the enterprise's revenue increased by 20% year on year.