| Sumario: | This report documents the scaling pathway of SIMCAST, a decision support system (DSS) for potato late blight (Phytophthora infestans), designed to translate weather-driven disease-risk modeling into actionable fungicide-timing recommendations that reduce avoidable yield losses and unnecessary pesticide use. Implemented with partners across Peru, Ecuador, and India, the initiative combined capacity building and user feedback with major technical modernization: the Lateblight Advisor web app was upgraded from an initial R/Shiny prototype to Shiny for Python, enabling stronger integration with modern data and AI workflows and real-time weather inputs (WeatherAPI); a SIMCAST Model Context Protocol (MCP) was developed to expose standardized model tools for multi-channel deployment; and GeoSIMCAST extended point-based modeling to national-scale risk mapping using gridded climate products and reconstructed dew point temperature (Tdew) to support humidity-driven risk estimation. Evidence from 30 farmer demonstration trials (2025) with resistant cultivars (Pollera and Watia) indicates that SIMCAST-guided scheduling can reduce disease intensity and optimize spray frequency, while yield responses warrant continued evaluation under broader field conditions. Together, these advances position SIMCAST as an interoperable, scalable digital advisory service that supports smarter fungicide use—right product, right timing, only when needed—and provide a clear strategy for wider adoption through phased scaling, partnerships, and extension-oriented delivery.
|
|---|