On September 14, the harvest began for a thousand mu of high-standard farmland in Yongfeng Village, Taihe Town, Dongpo District, Meishan City, Sichuan Province. Unlike previous years, this season marked the first application of a specialized AI model for rice field management across a 300-mu pilot area in the village.
Agricultural experts and growers gathered along the field ridges to observe the on-site yield measurement conducted by an expert group organized by the Sichuan Provincial Department of Science and Technology. The evaluation targeted the "Chengdu Plain Rice-Vegetable (Medicinal) High-Quality, High-Yield, and Efficient Production Technology Integration and Demonstration" project, which is being led by Sichuan Agricultural University.
Farming is now shifting from "experience-based field inspection" to "data-driven decision-making." The AI system collects data through drone patrols of the fields, providing farmers with precise guidance on rice transplanting, water and fertilizer regulation, and pest and disease warnings. "In the past, we relied on experience to inspect fields, but now drones do the patrols and the AI system pushes timely pest alerts, allowing us to take targeted action. Farming has become much more convenient," said Zhao Youyong, a major grain grower in Yongfeng Village, Taihe Town, Dongpo District.
According to the expert group's standardized yield measurement, all three core varieties in the 300-mu AI pilot field performed well. The high-quality hybrid rice "Huazheyou 210" achieved a measured yield of 826.8 kilograms per mu, the super-high-yield indica-japonica hybrid rice "Shengliangyou 222" reached 863.6 kilograms per mu, and the super-high-yield hybrid indica rice "Quanyou 169" produced 858.8 kilograms per mu. "The measured data proves that combining superior varieties, sound cultivation methods, and AI precision management can effectively unlock the yield potential of rice, providing a replicable technical pathway for large-scale yield improvement," said Ma Jun, a rice cultivation expert from Sichuan Agricultural University.
Ma Jun noted that the more than 240 new varieties displayed in Yongfeng Village this year showed excellent performance in yield, plant and leaf morphology, and grain quality. The application of smart precision drill seeding technology and AI has already shown preliminary results. The village's "smart farming" initiative did not happen overnight. In recent years, the Rice Research Institute of Sichuan Agricultural University has partnered with the Meishan City and Dongpo District Bureau of Agriculture and Rural Affairs, as well as the Minjiang Modern Agriculture Demonstration Park. Leveraging the Yongfeng Village rice new variety and technology pilot base, they have conducted pilot evaluations of over 670 new rice varieties and more than 20 new technologies, ultimately selecting 24 mechanization-friendly high-quality hybrid rice varieties and 5 high-yield, green, and efficient new technologies.
"In just three years, the yield of this demonstration field has increased by 152.6 kilograms per mu compared to the 711 kilograms recorded in 2023," said Yang Zhiyuan, deputy director of the Sichuan Provincial Rice New Variety and Technology Pilot R&D Platform. The related technologies are now being extended from Yongfeng Village to Chongzhou, Xindu, and other areas in Chengdu, helping to bridge the "last mile" in rice innovation, pilot testing, promotion, and application. Dongpo District in Meishan is actively advancing the transformation of rice production from an experience-based to a smart, intelligent model.