Design of a Low-Power Wireless Sensor Network for Soil Nutrient Monitoring in Vegetable Farms in Plateau State

Vegetable farming on the Jos Plateau in Plateau State, North Central Nigeria, depends on soil nutrient management, but soil testing is typically infrequent and costly, contributing to suboptimal fertiliser application and crop yield losses. This project designed a low-power wireless sensor network for continuous soil nutrient monitoring in vegetable farm plots. Five sensor nodes were deployed at 10-metre intervals across a demonstration plot, each equipped with NPK soil sensors, a DS18B20 soil temperature sensor, and a capacitive soil moisture sensor connected to a LoRa-enabled Arduino Pro Mini microcontroller. Sensor readings were transmitted at 15-minute intervals using LoRaWAN protocol at 433 MHz to a gateway node connected to a Raspberry Pi base station. A ThingSpeak IoT dashboard displayed real-time and historical nutrient, temperature, and moisture data with fertiliser recommendation logic implemented as Python scripts. The system was deployed and operated for 30 days on a 0.5-hectare tomato plot near Vom. NPK sensor calibration against laboratory wet chemistry analysis showed correlation coefficients of 0.84 for nitrogen, 0.79 for phosphorus, and 0.86 for potassium. LoRa communication range was verified at 420 metres in farm-field conditions. Node battery life exceeded 45 days on two AA lithium cells at the selected duty cycle. The study recommends combining soil nutrient data with yield harvest records to train a machine learning recommendation model for Plateau State vegetable crop types.

Keywords: wireless sensor network, soil nutrient, LoRaWAN, precision agriculture, Plateau State