Smart farming practices are inspired by sensing technologies, the main variables to be measured are soil and climate. The collection and processing of these data from different crops leads to better decisions that can benefit the entire value chain of planting and harvesting of products. For banana crops, the scope of this research, an architecture was designed to collect and transmit data regarding climate and soil variables with the greatest impact on the growth, maintenance and harvesting of the crop. These data are stored in a NoSQL database ideal for working with large volumes of data and displayed to the end user through an web application developed under Node.js. The application also allows the approximate calculation of the degree days to estimate the harvest time of the crop. This architecture was planned taking into account the main energy and communication obstacles that exist in the rural areas of the Piura region in northern Peru.