Enhancing Meghdoot: Integrating AI for Smarter Agricultural Advisories

Vista sencilla de ítem
dc.creatorDhulipala, Ram
dc.creatorSingh, Kanika
dc.date2024-12-29
dc.date2025-01-31T09:34:44Z
dc.date2025-01-31T09:34:44Z
dc.date.accessioned2026-06-27T16:48:49Z
dc.descriptionDigital Innovation Initiative at ILRI, in collaboration with partners, is integrating Artificial Intelligence (AI) into Meghdoot to enhance its efficiency and accuracy. A pilot project has tested AI models, such as Random Forest regression, Naive Bayesian, and Stacked Models, alongside OpenAI prompt engineering. Conducted at three locations in India, the pilot has demonstrated promising results. Efforts are underway to refine machine learning models, incorporate expert knowledge, and explore techniques like noisy labels to improve advisory quality. A web-based platform has also been developed to automate advisory generation, allowing users to select parameters like location, crop type, and AI model. The system generates personalized advisories using historical, observed, and forecasted weather data. It provides both AI-generated and traditional advisories, along with weather forecasts and SMS summaries for easy dissemination. Moving forward, the goal is to integrate this AI-powered advisory system into Meghdoot, scaling it nationwide to improve agricultural decision-making, enhance sustainability, and increase resilience among farmers.
dc.formatapplication/pdf
dc.identifierhttps://hdl.handle.net/10568/172636
dc.identifier.urihttp://hdl.handle.net/123456789/133895
dc.languageen
dc.publisherInternational Livestock Research Institute
dc.rightsOpen Access
dc.sourceDhulipala, R. and Singh, K.2024. Enhancing Meghdoot: Integrating AI for Smarter Agricultural Advisories. Progress Report. Nairobi, Kenya: ILRI.
dc.subjectagriculture
dc.subjectclimate change
dc.subjectfood security
dc.titleEnhancing Meghdoot: Integrating AI for Smarter Agricultural Advisories
dc.typeReport

Archivos

Colecciones

Instituto Internacional de Investigaciones Pecuarias (ILRI)