Digitalization of agriculture: application of IoT and machine learning for enhanced crop management

Abstract

Amid climate change and rising global food demand, digital technologies are emerging as a critical enabler of sustainable agricultural development. This study explores the integration of Internet of Things (IoT) and machine learning techniques into crop management systems. Research Objective: To develop and test a prototype of an intelligent agro-monitoring system capable of predicting the need for irrigation and fertilizer application based on data from soil sensors, weather stations, and satellite imagery. A field experiment was conducted in 2024 on a 5-hectare plot in the Stavropol Region, collecting data on soil moisture, air temperature, and the Normalized Difference Vegetation Index (NDVI). A Random Forest algorithm achieved 92 % accuracy in irrigation demand prediction. The results demonstrate an 18 % reduction in water consumption and a 12 % increase in wheat yield, confirming the practical applicability of the proposed system for small- and medium-scale farms.