GRENZE International Journal of Engineering and Technology
Vol. 12
(2026), Issue 2
Agri-Route: A Smart Advisory System for Farmers and Transporters
Authors
Milind Rane, Varad Pakhale, Medha Wyawahare, Vaibhav Panchal, Saransh Nakhale
Abstract
Agri-Route is a production-to-production multi-agent Human-AI voice-first advisory platform that aims to transform support decisions for smallholder farmers. It is composed of a crop planning agent, a disease detection agent, a weather forecasting agent, a market price analysis agent, a financial consultancy agent, an input cost optimisation agent, and a perishable logistics agent, all running under a unified cloud-native coordinator-orchestrator architecture. Based on Next.js 15, React 19, the OpenAI GPT-4o Realtime API, and Firebase, Agri-Route offers natural-language voice interaction in Hindi and Marathi and achieves an average median end-to-end latency of 680 ms. A fresh approach, incorporating the Mapbox geospatial routing for a produce specific exponential decay model with live OpenWeatherMap data, is proposed in the Freshness Score Calculator to reduce loss after harvesting during transit. A controlled user study with 45 smallholder farmers in Pune district, Maharashtra showed a significant 34.6 percentage-point increase in task-completion accuracy (p < 0.01) and a 40% reduction in decision-time over traditional extension-service baselines. System Usability Score was 74.2, which was higher than the industry-standard 68-point threshold.
Pages:
6823 - 6830