This work shows the modeling, identification and control of a cascade tank system. The system was built to scale in order to carry out the experimental part of the project on a real system. Initially, a data collection was carried out with which a model based on neural networks was obtained to identify the dynamic behavior of the system. Subsequently, using the same data collection, a neural network was trained to control this system through a reverse neuronal control scheme. Both neural networks were implemented in the Arduino platform with which the real tank system could be emulated and controlled. The above allows to show the versatility of said platform for this type of applications achieving low cost implementations.