فهرست نشریات علمی دسترسی آزاد ایران

Artificial intelligence (AI) is becoming increasingly pivotal as a foundational component of precision and sustainable agriculture, where sensing, connectivity, and data-driven automation work together to reduce uncertainty in biological production systems, optimize inputs and resource use, and mitigate environmental impacts. Despite increasing progress, the evidence remains scattered across some disciplines such as computer vision, remote sensing, agricultural engineering, and operations research, often relying on heterogeneous datasets and inconsistent validation protocols. This review compiles peer-reviewed studies on AI applications in agriculture, especially in crop production and protection, livestock monitoring, and post-harvest decision-making and supply-chain management. Using PRISMA 2020 reporting guidelines, we analyzed a carefully curated set of DOI-indexed publications (n = 141) and their citations to identify influential methodological contributions. Furthermore, we provide a critical evaluation of the methodological rigor of the published articles on AI-powered agriculture, ranging from precision agriculture, crop observation and disease detection, yield and resource-efficiency estimation, to autonomous agricultural machinery, and agricultural logistics optimization. We examined how data types and task structure influence model choice, for instance, highlighting the predominance of deep convolutional and transformer-based architectures for imaging and diagnostics, tree ensemble methods for tabular agro-climatic prediction, and multimodal pipelines that integrate satellite imagery, soil properties, and farm management records to forecast yield and irrigation needs. Across these areas, the review emphasizes recurring gaps between benchmark accuracy and real-world robustness, driven by domain shifts, limited labeled data, and the risk of dataset leakage. Consequently, to create trustworthy and deployable systems, we highlight the emerging importance of external validation, measures of uncertainty, explainable AI, and semantic tools for model interpretability, as well as energy-efficient “green AI” training and edge deployment. In the following, we explore the potential of the emerging convergence of AI and nanoagriculture as a promising strategy towards next-generation sustainable agricultural systems, given the growing need to harness technologies to address challenges in the agricultural sector, including the issues posed by climate change and the resulting environmental concerns. AI-driven modeling and optimization enable the rational design of smart and safe application of nanomaterials for agriculture, including nanofertilizers, nanopesticides, and nanosensors. These approaches offer the potential to separate productivity gains from input intensity by minimizing environmental emissions, improving delivery efficiency, and increasing spatiotemporal precision. In addition, some challenges, such as data scarcity and safety, were discussed. By addressing technological potential and systemic limitations, this review provides a balanced framework for advancing AI-enabled nanoagriculture in support of resilient, environmentally friendly, and sustainable food systems.