Soybean ranks among the most economically significant commodity crops globally, driving multibillion dollar agricultural supply chains across the United States, India, Brazil, and beyond. Despite its mass production scale, soybean operations suffer from inefficient disease management practices where conventional broadcast pesticide application exposes entire fields to unnecessary chemical inputs regardless of actual infection status, generating measurable supply chain losses and inflating operational costs across the commodity value chain. Structured against the Supply Chain Operations Reference (SCOR) model’s Plan, Source, Make, Deliver, and Return processes, this study introduces Fasal AI Dost, a farmer centric AI framework delivering intelligent crop disease detection directly into soybean farming operations to strengthen supply chain performance from the field level upward across all five SCOR process domains. A systematic review following PRISMA guidelines analyzing 127 peer reviewed articles (2018 to 2025) reveals that fuzzy logic approaches, representing only 9% of existing AI agricultural solutions, offer 90 to 95% computational efficiency advantages over dominant deep learning frameworks (68% of studies) while remaining deployable without specialized infrastructure. Existing studies combining SCOR and fuzzy logic confirm the methodological viability of this integration across supply chain performance evaluation contexts (Ganga and Carpinetti, 2011; Lima Junior and Carpinetti, 2016; Nilashi, 2025), yet no prior study applies this combination to field level soybean disease detection across geographically contrasting production regions. Building on this evidence, SOYFLEX (Soybean Leaf Fuzzy Expert Classifier) is pilot tested using field data from Maharashtra, India and South Dakota, USA, two major soybean production regions representing diverse climate conditions and operational contexts within global soybean supply chains.

The pilot classifier achieves 78.4% accuracy and 0.893 AUC ROC using five interpretable visual features operable on standard smartphone cameras without internet connectivity. Precision targeted pesticide application projects $75 to $150 savings per hectare and 50 to 80% chemical input reduction, translating to an estimated $400 to $600 million in annual supply chain savings at combined regional scale. Mapped against SCOR performance metrics, SOYFLEX contributes to Perfect Order Fulfillment through reduced disease driven yield loss, Supply Chain Flexibility through farmer deployed field intelligence at the point of infection onset, and Cost of Goods Sold reduction through precision input targeting that eliminates unnecessary broadcast pesticide application across healthy field areas. This research establishes an adoption centric design paradigm demonstrating that soybean supply chain efficiency gains at global commodity scale originate at the field level, in the hands of the farmer who grows the crop, and that accessible, interpretable AI is not a compromise on technical rigor but the most direct path to real world supply chain performance improvement.