A Relational Database Management System approach for Data Integration in manufacturing process

In this paper we introduce a data integration system by implementing a function into the context of PostgreSQL. The aim of this work is to collect files to process from two different data sources (a platform of Physical Testing Software (PTS) and another one of Physical Simulation Software (PSS)), in order to retrieve specific records through a query and integrate them. Both these platforms contain a large amount of files in semi-structured or unstructured format. This approach allows analysing data from different sources and creating a database remaining always in the PostgresSQL context. Indeed, the code is modular and it can be customizable for the specific scope. This approach reduces the information exchange with the client and the computational cost, because the instructions are applied all in once (on-the-fly) to the files stored in the Network File System (NFS). Furthermore, the integration outputs and the related data can be stored within it. One of the objectives of this work is to perform a new product introduction (NPI) for multi-sources data retrieval and integration. The result is a modular approach, customizable and suitable for the most of the integration and retrieval issues. The architecture proposed allows all the authorized users to access to the data in parallel, independently and on the same device, by running query straightly on the file through the Structured Query Language (SQL).