An Architecture for Efficient Integration and Harmonization of Heterogeneous, Distributed Data Sources Enabling Big Data Analytics

摘要

We present a lightweight integration architecture as an enabler for the application of process optimization via Big Data analytics and machine learning in large scale, multi-site manufacturing companies by harmonizing heterogeneous data sources. The reference implementation of the architecture is entirely based on open-source software and makes use of message queuing techniques in combination with Big Data related storage and extraction technologies. The approach specifically targets challenges related to different network zones and security levels in enterprise information architectures and across divergent production sites.