Sustainable Plant Disease Management with Real-Time Crop Optimization

Abstract

Plant diseases significantly threaten global food security, often leading to severe yield losses and unsustainable reliance on chemical usage and pesticides. This paper presents an integrated, real-time system for sustainable plant disease management using Internet of Things (IoT) sensors, deep learning models, and cloud-edge computing. The proposed framework enables early disease detection and adaptive crop optimization by fusing environmental telemetry with AI-driven image diagnostics. Using the PlantVillage dataset and real-world sensor data, the system achieves 99.1% disease detection accuracy, a 27% reduction in pesticide usage, and a 22% improvement in crop yield, a critical metric in assessing the broader effectiveness of plant disease management strategies compared to leading benchmarks. Field trials confirm its efficacy in enhancing farm productivity while minimizing environmental impact. This work demonstrates a practical, scalable solution for precision agriculture that aligns with the principles of sustainability, resilience, and data-driven decision-making.

Plant phenotyping relevance

植物画像から病害状態を推定するAI診断とセンサー統合基盤が研究の中心であり、植物病害フェノタイプの実質的な取得・評価を行っている。

Code and data availability

The paper's disease-classification measurements are based on the public PlantVillage dataset, cited with an explicit Kaggle URL. The real-world IoT sensor/field-trial data and the authors' code or trained MobileNetV2 model have no stated public availability.