The present work consists of carrying out the identification and implementation of a predictive controller in a thermal system to follow a didactic design methodology for learning in systems identification, analysis, and controller design. For the identification of the thermal system, the method of least squares was chosen because it allows to obtain a faithful mathematical model that describes the temperature system. The data used in the system identification step are from a didactic industrial plant of a temperature system. From the model obtained for the system in question, the design and implementation of a classic Proportional Integral Derivative (PID) controller and a predictive controller are carried out to verify the behavior of the system against the actions of both controllers. The entire implementation will be carried out in the MATLAB/Simulink computer simulation environment.