Zhinong AI: A Design-Science Study of an AI-Enabled Agricultural Decision-Support Platform for Smallholder Production

The paper presents the Zhinong AI Agricultural Decision Platform, an integrated system designed to support smallholder farmers by combining various AI functionalities such as natural language processing for question answering, image-based crop disease diagnosis, and workflow orchestration. It introduces a layered system architecture that encompasses a closed-loop decision process, including sensing, analysis, planning, execution, and feedback, along with a governance framework addressing data provenance and model risk. This study is significant as it provides a structured research framework for developing AI agricultural prototypes into accountable decision-support systems, although it does not present quantitative performance metrics due to the lack of field data at the time of writing.