Artificial Neural Networks (ANNs) were used in this present research work to predict electricity consumption of manufacturing machines based on data collected from manufacturing industry which deals with the production of medical replacement. Data related to production and electricity consumption were collected for a period of six months from four different machines in a production line. Data related to electricity were collected every 15 min at sampling rates. While the data related to production rate was irregular and registered by the technical in charge of the machine. Artificial Neural Networks (ANNs) such as NNARX, NNARMAX and NNOE nonlinear neural network were used, and their accuracy was verified by utilizing criteria such as Mean Squared Error, Mean Absolute Error and Goodness of Fit. The data was collected to develop and then predict electricity consumption and production of machines up to 24 h ahead. The main objective of this research is to build a methodology that allows to build models based on inputs and outputs for the prediction of electricity consumption throughout the Value Stream Line (VSL) composed by four machines. Finally, results analysis presents the methodology composed by six steps where firstly we start with the data analysis related to electricity and production, secondly development of ANNs models, thirdly training of ANNs models, fourthly validation of ANNs models, fifthly model analysis on different step ahead prediction, and lastly selection of the best models.