Education + Industry 4.0: Developing a Web Platform for the Management and Inference of Information Based on Machine Learning for a Hydrogen Production Biorefinery

摘要

Nowadays, humans depend on the operation of different complex systems. For example, transportation systems (trains, ships and planes), public services (water, gas and electricity), manufacturing-plants, hospitals and banks, to name a few. In recent years, given the improvement of sensors it is normal to have a large amount of data of a complex system. This creates a major challenge for analysis, visualization, and decision making about this data. In these days, it is not enough a statistical analysis of the data, but a machine learning needs to be applied for better inferential information. In this paper, as part of our formation in masters in computer science, we propose a new architecture for a web platform for the management and inference of information based on Machine Learning. It can receive, clean, pre-process and transform the data for the statistical analysis and application of machine learning algorithms. The proposed architecture is validated with experimental tests obtained from a simulator of a hydrogen production biorefinery, but can be abstracted and applied in different complex systems.