On August 24, MAP, a subsidiary of Sinochem Holdings, deployed its large-scale agricultural planting model called iMAP across China, reaching over 2,500 agronomists. The model combines crop science, a large-scale model, and intelligent agents into a unified system, creating a one-stop AI solution for farming. Already applied on over 1 million mu of farmland, it helps cut water and fertilizer use by 10 to 20 percent and boost yields by 8 to 10 percent, able to effectively foster new quality productive forces in agriculture.

Leveraging business units that span the entire agricultural industry chain and specialized agricultural technology service platforms across 26 provinces, autonomous regions and municipalities, MAP integrates firsthand data, works extensively with universities, research institutes, industry organizations, and technology companies, and consolidates a vast agricultural knowledge base exceeding ten million entries. These efforts have generated 6 thematic data domains and over 10 multimodal, high-quality agricultural datasets for farming.

The iMAP model is built on an innovative algorithmic framework that is explainable, traceable, and rooted in crop science. By simulating 5 key processes of crop growth, it introduces a new mechanism-driven approach to AI in agriculture. The system supports 8 intelligent applications covering the full farming cycle—from soil preparation and planting to crop management and harvesting—providing a one-stop AI solution while reducing reliance on experience in agricultural production.

Looking ahead, MAP plans to open up the iMAP model to the public and accelerate its commercial rollout. Positioned as a new farming tool for the digital era, the model is expected to play a key role in building up China’s strength in agriculture and driving forward the implementation of the AI development strategy.