This document describes the process of acquiring environmental data that feeds an intelligent irrigation control system which, based on the calculation of the evapotranspiration of a crop, manages to calculate its water needs by means of artificial intelligence, with a method derived from random forest to supply them. After considering the problem of irrigation, and the justification of a solution based on the Internet of Things (IoT) as satisfactory, the variables involved in the process and the characteristics of the data produced by the sensors are specified; The data capture process is developed on an IoT architecture based on knowledge management with the phases of: sensing, communication and analytics, referring to the R software components that have been implemented to carry out this process, culminating with the analytical projections of the irrigation. This process is then emulated by an artificial intelligence which is chosen among 20 different topologies to choose the best one based on the reduction of the mean square error and emulate the irrigation estimation by means of an AI. As irrigation is the main aspect of crop performance, it is concluded that an inherent need in the field sector is met - which is not yet automated - proposing a solution for the specific conditions of the Colombian field.