Big data analytics for intra-logistics process planning in the automotive sector

Abstract

The manufacturing sector is facing an important stage with Industry 4.0. This paradigm shift impulses companies to embrace innovative technologies and to pursuit near-zero fault, near real-time reactivity, better traceability, and more predictability, while working to achieve cheaper product customization. The scenario presented addresses multiple intra-logistic processes of the automotive factory Volkswagen Autoeuropa, where different situations need to be addressed. The main obstacle is the absence of harmonized and integrated data flows between all stages of the intra-logistic process which leads to inefficiencies. The existence of data silos is heavily contributing to this situation, which makes the planning of intra-logistics processes a challenge. The objective of the work presented here, is to integrate big data and machine learning technologies over data generated by the several manufacturing systems present, and thus support the management and optimisation of warehouse, parts transportation, sequencing and point-of-fit areas. This will support the creation of a digital twin of the intra-logistics processes. Still, the end goal is to employ deep learning techniques to achieve predictive capabilities, all together with simulation, in order to optimize processes planning and equipment efficiency. The work presented on this thesis, is aligned with the European project BOOST 4.0, with the objective to drive big data technologies in manufacturing domain, focusing on the automotive use-case.