A Low-Cost Autonomous Agricultural Rover for GPSEnabled Precision Fertilizer and Soil Nutrient Management

Small and medium farmers do not have much access to precision agriculture because of the expensive costs of the system and significant infrastructure needs. This paper will describe a low cost autonomous agricultural rover, which is capable of carrying out site specific soil nutrient analysis, and control fertilisers, using GPS. The suggested system combines the sensing of soil electrical conductivity (EC), soil moisture, GPS localization, micro-weather, and lightweight vision-based weed detection using an ESP32-based platform. The potentiation of the spatial soil data is carried out by Inverse Distance Weighting (IDW) interpolation and K-means (K = 3) clustering to develop nutrient management zones, and the fertilizer recommendation is derived by use of Nutrient-Per-Water (NPW) index. Soil mapping in field experiments on a 0.5- hectare maize plot shows an accuracy approach of 5 percent, soil moisture correlation of R 2 = 0.87 as well as a maximum of 38 percent of lowering of the use of fertilizer without a noticeable decrease in yield. The rover consumes a mean power of about 32 W, and can be continuously operated 3.5 hours, and the total system costs less than USD 50. The findings reveal that the system proposed has a feasible and scalable answer to precision fertilization in easily accessible farming conditions.