This paper addresses inefficient irrigation caused by spatial variability when a single sensor is used for an entire field. Our proposal is an IoT-based precision irrigation system whose area is a 100 m2 field that is subdivided into six equal parts and measured by ESP32 nodes, which have soil-moisture probe and DHT11 temperature/humidity sensors. Nodes send readings to an ESP32 over MQTT; the ESP32 calculates an average of soil-moisture, temperature and humidity and sends them to a Raspberry Pi 4 serving a HiveMQ broker and a Node-RED dashboard. The dashboard is real time visualized and controlled by pump, and Pagekite opens the dashboard to be accessed remotely safely. It was assessed according to the quantifiable parameters (soil moisture, temperature, humidity, averaged field values) and system performance (real-time data transmission, the responsiveness of the dashboard and automatic triggering of pumps when moisture levels were exceeded). Experimental implementation showed that there were consistent real-time data acquisition, real-time visualisation of per-node and averaged values and automatic control of the pumps through the set moisture limits.