AgriAdvisor: On-Device Embedded LLM System for Crop Disease Detection and Prescriptive Advisory with Differential Privacy

Read full text Summarise with VIRAT
Cited by: 0
References: 7

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Current crop disease detection systems primarily focus on image classification but lack actionable, contextsensitive advisory information.Moreover, they often rely on cloud servers, compromising privacy and requiring stable internet, which is challenging in rural areas.This paper introduces AgriAdvisor, an on-device embedded system integrating a lightweight disease detection model with a fine-tuned small Large Language Model (LLM) for prescriptive advisories, incorporating differential privacy to safeguard user data.Deployed on mobile devices, AgriAdvisor detects diseases from field images and provides localized treatment suggestions based on crop type, growth stage, and climate.Evaluations on multiple crops in Nashik, India, demonstrate high detection accuracy (95% F1-score), useful advisories (rated 4.5/5 by farmers), low latency (under 2 seconds), and preserved privacy.This novel integration addresses gaps in existing literature by combining detection, advisory, privacy, and edge computing.We expand on prior works by incorporating multi-modal data handling and federated learning enhancements, drawing from recent advancements in deep learning for agriculture.

Fields of Study & Subjects

Smart Agriculture and AI Advanced Neural Network Applications IoT and Edge/Fog Computing

Publication details

Journal / source
International Research Journal of Modernization in Engineering Technology and Science
Publication date
30 March 2026
Language
EN