Apache Spark for Digitalization, Analysis and Optimization of Discrete Manufacturing Processes

Abstract

Digitalization, analysis and optimization of discrete manufacturing processes represent a research challenge because the data generated by the sensors that monitor the manufacturing processes are characterized by noise, missing values, many features due to the fact that the assembly processes involve many different assembly steps and often the resources that are used in the manufacturing processes are applied inefficiently. Digitalization of assembly processes using the latest technologies, analysis of data generated by the monitoring sensors using big data technologies and optimization of the manufacturing processes by identifying the steps that have the highest impact on the final output represent serious research challenges, and in this chapter, we approach these challenges by: (1) creating a platform in which the users can visualize the assembly steps of the discrete manufacturing processes, (2) identifying the products that have manufacturing faults using machine learning and deep learning models and (3) optimizing the parameters of the machine learning and deep learning models using bio-inspired heuristics.