An AI-Driven End-to-End Agricultural Guidance System with Multilingual and Voice Support

Abstract

Farmers have a need for integrated, intelligent decision support systems to help improve productivity amidst challenges posed by increasing climate variability, increasing resistance of pests to existing control methods and decreasing price stability. To meet this need, an end-to-end agricultural guidance system (an integrated AI-based digital platform) has been developed that provides support throughout the entire crop lifecycle. The end-to-end agricultural guidance system will include eight different modules: Crop Types (using Machine Learning); Plant Disease Detection (using Convolutional Neural Networks); Classifying Disease Severity; Proactive Disease Risk Prediction; Community Based Regional Disease Alerting System; Weather Analytics Dashboard; Natural Language Processing (NLP) Based Multilingual Voice-Enabled Chatbot; and Post-Harvest Price Analysis Engine. The central difference between the end-to-end agricultural guidance system and the previously utilized reactive systems is that rather than only addressing one isolated agricultural challenge at a time, the end-to-end agricultural guidance system operates from a proactive and holistic perspective. Experiments conducted on each of the eight modular components of the end-to-end agricultural guidance system have demonstrated the potential to provide accurate results and to provide farmers with actionable real-time advice regardless of their language or level of education. The end-to-end agricultural guidance system is intended to assist farmers in growing a sustainable, data driven, and farmer centric agriculture.