Optimization and estimation framework of smart farm based on spatial data mining and geostatistics

The development of integrated information systems in the field of rice plant has become an urgent need for policy makers at both provincial and national levels, specialized in developing countries such as Indonesia. At this time, data related to rice avaibility is still spread across several agencies and difficult to access easily and quickly to support strategic and rapid decision making. This data is the level of fertility of rice plants, fertilizers and water content at certain locations and times. Fertilizers, new superior varieties and water content are production factors that are vital in supporting the efforts to increase national rice production. New superior varieties including hybrid rice are generally responsive to macro fertilizers NPC (Nitrogen-Phosporus-Calcium), where fertilization efficiency and effectiveness is very dependent on local location. The use of fertilizers with excessive doses must be prevented by socializing the right dosage specifically for location and time. The use of fertilizers with high doses of nitrogen also results in plants being more susceptible to plant pests. Based on the consideration of efficiency and sustainability, the use of uniform recommended doses for large areas and not considering the specific conditions of the plant is no longer relevant to be applied. Therefore the distribution of the characteristics of rice and soil at specific locations and times is very important information. Research objectives to develop technology that integrates operational information systems at farmer levels (SMS, GSM or IoT) and information systems in management levels with GIS, Spatial data mining, Kriging Interpolation and Artificial Intelligence systems that have prediction ability and optimization of food security problems, especially rice. Spatial data mining with approached geostatistics used to map the distribution of various factors, Nitrogen, Phosphousr, Calcium and Water content that influence the growth rate of rice plants. The results of this study are a smart farm framework that can estimate and optimize sustainability and availability of rice.