In arid and semi-arid regions, water scarcity is one of the most significant challenges in agriculture. Efficient water management is important where precipitation is low, and groundwater resources are also depleting. Most existing systems rely on a single sensor input, require continuous internet connectivity, and offer limited adaptability to changing field conditions. This research presents a Fuzzy-IoT irrigation system, implemented for the GC-2 variety of castor in Bhuj, India. The system combines data from multiple sensors, including soil moisture, rainfall, temperature, and relative humidity. The system consists of a sensor node and a collector node. The sensor node measures field parameters and sends them to the collector node. The collector node performs local fuzzy-logic decision-making to control the pump without an internet connection and uploads data to the ThingSpeak cloud via an ESP-based module for monitoring. In this system, data are applied to a fuzzy inference framework to enhance irrigation scheduling and improve decision-making in semi-arid regions. Moreover, to ensure practical usability in rural areas, the framework also utilizes LoRa, which provides infrastructure-less communication and long-range operation without continuous internet access. Field evaluation revealed that the system reduced the total irrigation water requirement to 337.91 m3/acre, achieving 26% water savings. Specifically, this performance in water saving is higher than the other state-of-the-art fuzzy-based irrigation approaches. Overall, the results demonstrate that fuzzy IoT irrigation has delivered sustainable water savings compared with traditional irrigation practices in semi-arid agricultural regions.
Vahora et al. (Sun,) studied this question.