Sichuan Agricultural University and Dekang Launch Swine Industry AI Platform

Sichuan Agricultural University and Sichuan Dekang Agriculture and Animal Husbandry Food Group signed a strategic agreement on August 28 to co-develop a vertical foundation model and intelligent agent platform for the swine industry, according to People's Daily. Designed to address data silos, manual labor reliance, and genetic selection challenges, the initiative targets the entire pork production chain. Initial deployments will focus on core breeding centers, commercial pig farms, and slaughter facilities before expanding into supply chain and food safety management.

The collaboration illustrates how generative AI and multi-agent systems are migrating into heavy primary industries. Pork production remains a cornerstone of China's agricultural economy, and integrating vertical models into breeding and farm management could help institutionalize tacit veterinary knowledge, lower operational overhead, and reduce mortality rates across large-scale facilities.

Sichuan Agricultural University and livestock enterprise Sichuan Dekang Agriculture and Animal Husbandry Food Group signed a strategic partnership agreement on August 28 to co-develop an artificial intelligence foundation model and intelligent agent platform for the swine industry, according to People's Daily. The initiative seeks to integrate artificial intelligence across the entire swine production lifecycle, establishing a technical framework spanning data governance, knowledge structuring, model development, agent coordination, system integration, and iterative validation. By combining Dekang's industrial operational data and production scenarios with the university's research in animal genetics, nutrition, and agricultural computing, the partners aim to tackle long-standing pain points including fragmented data silos, heavy dependence on manual labor, and imprecise genetic selection. According to the project team, initial validation will prioritize three core operational environments: breeding facilities, large-scale commercial pig farms, and slaughterhouses. Core breeding stations will deploy the system for automated phenotype collection, genetic evaluation, genomic selection, and mating optimization. Commercial production farms will use intelligent agents for health management, precision feeding schedules, production dispatching, and early anomaly warnings. At the slaughter stage, carcass and meat quality measurements will be recorded and channeled back into upstream breeding and husbandry models to refine selection criteria. Later stages are planned to extend into food quality control, supply chain tracking, and business analytics. Zhu Li, dean of the College of Animal Science and Technology at Sichuan Agricultural University, stated that AI integration allows institutional knowledge and specialized expertise accumulated by senior researchers to be codified and scaled via algorithms across commercial sites. A Dekang representative noted that testing will center on real-world production metrics, verifying whether automated data collection functions consistently and whether analytical outputs translate into actionable barn-level interventions through human review. Under the framework agreement, the university and Dekang will also collaborate on patent applications, software copyright filings, industry standards development, and interdisciplinary talent cultivation to create a standardized operational template for livestock digitalization.