European Journal of Sustainable Agroecosystems

Global agriculture is facing a major challenge: irrigation already accounts for the major share of freshwater withdrawal around the world, but it is characterized by inefficient water and fertilizer use, which contributes to depletion of the resources and deterioration of soil and water quality. The Internet of Agricultural Things (IoAT), which is the agricultural branch of the Internet of Things that consists of soil and nature sensor systems, wireless communication, edge computing, cloud computing and AI systems, has appeared as the key concept in solving this problem based on smart irrigation and mechanization of feeding. The article reviews the existing research focusing on smart irrigation and mechanization of feeding based on the IoAT concept. All research results analyzed show that moisture, electrics and pH sensor systems in conjunction with automatics of valves and fertilizer injecting devices allow to apply water and fertilizers directly on the basis of current requirements and not in compliance with a schedule. The results show that WUE of [amount] and NUE of [amount] are achieved at various configurations based on multiple use of fertigation and irrigation methods that provide [amount] improvement in yield. Wireless sensor LoRaWAN and similar low-power, wide-area network types of technology combined with edge devices powered by solar energy have been chosen as preferred means of communication for deployment of IoAT on-field as they provide optimal balance between long-distance communication, energy efficiency and price. AI-based support, usually using some machine learning models for prediction or classification purposes turns multi-sensor observations into specific plans for irrigation and fertilization applications, often showing high accuracy above 90% in controlled conditions. The review also summarizes existing problems, including high cost of sensors and other devices, requirement of constant calibration and maintenance for EC/pH/nutrient sensors, problems with interoperability of devices of different manufacturers, as well as inconsistent information on operational characteristics of the systems like the response time and energy consumption. Thus, it concludes that implementation of IoAT based solutions in irrigation and fertilizing processes is a good scientifically-sound way towards sustainable water and nutrients use.