Toward Sustainable Farming: Artificial Intelligence Applications in the Detection and Management of Crop Diseases and Pests: A Systematic Review

Abstract

Crop production faces increasing threats from pests and diseases, which significantly reduce yields and compromise food security, a challenge further intensified by climate change. Conventional pest and disease management approaches are often labor-intensive, time-consuming, and environmentally unsustainable. In this context, Artificial Intelligence (AI), particularly Convolutional Neural Networks (CNNs), has emerged as a powerful tool for accurate and timely detection of crop pests and diseases. This study presents a systematic review of recent advances in CNN-based approaches for pest and disease detection in agricultural systems. A comprehensive literature search was conducted across Scopus, IEEE Xplore, PubMed, Springer, and ScienceDirect for English-language studies published between 2019 and 2025, using predefined keywords related to AI, CNNs, crop diseases, and pest monitoring. Eligible studies were selected based on inclusion and exclusion criteria derived from the PICo framework. Extracted data included crop types, targeted pests and diseases, CNN architectures, dataset characteristics, and performance metrics such as accuracy, precision, and recall. A total of 100 studies were included and synthesized narratively. The findings indicate that CNN-based models consistently achieve high detection performance across a wide range of crops and agro-climatic conditions, with many studies reporting accuracy rates exceeding 95%. These AI-based approaches outperform traditional visual and manual inspection methods in terms of speed, precision, and scalability. Moreover, the integration of CNN models with IoT and edge computing platforms shows strong potential for real-time, field-deployable applications. However, challenges remain, particularly regarding data availability, computational requirements, and deployment in resource-constrained environments. Overall, CNN-based AI technologies represent a promising pathway toward sustainable crop protection, improved agricultural productivity, and enhanced food security.