Deployment of a Smart Irrigation Control System with Capacity-Based Moisture Sensors on a Production Farm

Abstract

This paper demonstrates a smart irrigation system prototype, based on Internet-of-Things (IoT) devices and cloud computing, in a fully operating environment. The system architecture was developed carefully with a special focus on robustness against both environmental and human external factors. The platform, which is deployed in the cloud and connected to the edge-layer via a bidirectional LoRa wireless network, was based on data gathering from the field using a set of cost-effective capacity-based moisture sensors. A hysteresis-based control structure implemented in the cloud send the control commands. The demonstration was performed on a strawberry production farm in Itaguá, Paraguay, during a 2-month period. Details of the implementation are provided, as well as an assessment of the irrigation system performance. It was found that the automated irrigation systems consumed slightly less water than manual irrigation by the farmer, but the efficiency of the automated system reached 91.4%, compared with 62.1% for manual irrigation. Finally, real-life issues encountered during the operation are discussed to illustrate the robustness of the prototype. In spite of these issues, the irrigation system was able to keep the moisture within the prescribed band most of the time, about 71.1% of all the samples, and 87.9% under normal operation.