The problem of the scarce of drinking water is increasing every day around the world and is required of new monitoring systems in order to improve the care and management of water. The intelligent sensors with function of data fusion are a good alternative to build intelligent monitoring systems. This paper presents a data fusion method based on least squares technique to design one intelligent sensor using a microcontroller of low computational capability. Three variables are used by the data fusion model to estimate the water consumption; the result is compared with a real value of water consumption. The intelligent sensor proposed is able to detect if the water consumption is normal or not and its performance was simulated using real data of water consumption of an Instituto Tecnológico de Chihuahua facility. Intelligent water consumption monitoring systems can be built for warning about improper water use and improve its care and management.