Amit Behera, Dhruv .
Precision agriculture needs irrigation management systems that are reliable and can be de-ployed in remote settings without much infrastructure. Presented in this paper is a multi-sensor Internet-of-Things (IoT)-based smart irrigation monitoring system, leveraging Arduino-based real-time control combined with Raspberry Pi edge computing and long-range LoRa communication. The proposed architecture monitors field conditions continuously by collecting environmental data from various sensors, including soil moisture, temperature, humidity, water level, and turbidity, and makes irrigation decisions based on a threshold-based control mechanism augmented by hysteresis protection. A 72-hour experimental deployment was carried out for system performance evalua-tion, during which, behavioral studies and environmental trends were analyzed based on the data recorded by the environmental sensors. The irrigation cycles are triggered automatically when the soil moisture exceeds the given threshold while other environmental parameters remain stable un-der monitoring. A classification using a Random Forest machine learning algorithm was developed for predicting irrigation requirements, with an overall classification accuracy of approximately 87% and an AUC value of just under 0.90. Communication performance tests exhibited a strong LoRa connectivity and high-performance transmission with Received Signal Strength Indicator (RSSI) values ranging from 33 dBm to 10 dBm and a packet delivery ratio of 98.5%. This architecture is predicated on deterministic microcontroller irrigation control integrated with edge-based analytics, limiting reliance on uninterrupted cloud connectivity yet allowing for monitoring agricultural ac-tivities in real-time. The experiment shows that the system is a scalable and cost-efficient solution to precision irrigation and resource-efficient water management in smart agricultural environments.