Geospatial Artificial Intelligence in Precision Agriculture: A Systematic Review of Applications, Multimodal Data Integration, and Decision Support

Abstract

Geospatial artificial intelligence (GeoAI) offers opportunities to translate heterogeneous agricultural observations into site-specific management decisions. This systematic review synthesizes literature published in 2019–2026, identified through Scopus, Web of Science, and supplementary searches, to examine applications, multimodal data integration, and decision support in precision agriculture. The narrative synthesis indicates a shift from isolated mapping and prediction toward integrated workflows combining satellite and UAV imagery, field sensors, weather, soil, and management records. Reported benefits include improved yield prediction, earlier stress and disease detection, more detailed soil mapping, and more targeted irrigation and input use. However, farming yield, environmental, and economic gains remain dependent on local conditions and operational validation. A central finding is that georeferenced data alone do not ensure genuinely geospatial AI: reliable workflows must address spatial dependence, scale mismatch, transferability, and uncertainty. Interpretability and integration into farm operations are equally important for actionable recommendations. Adoption remains constrained by data interoperability, connectivity, affordability, privacy, and technical capacity. Heterogeneous evidence and single-reviewer assessment limit interpretation. Future progress requires spatially informed, field-validated models, accessible decision support tools, and participatory governance to support sustainable and inclusive agricultural management.