A Framework For Adaptive Data Integration In Digital Production

Modern production processes' complexity increases steadily. Therefore, virtual planning has been prevailed as a method used to evaluate risks and costs before the concrete realization of production processes. In doing so, virtual planning uses a number of numerical simulation tools that differ in the simulated production techniques as well as in the considered problem domains. Users may choose between tailor-made, thus costly, simulation tools delivering accurate results and off-the-shelf, thus less costly, simulation tools causing post-processing efforts. Thereby, simulating a whole production process is often hardly realizable due to insufficient prediction accuracy or the missing support of a production technique. The supposed solution of interconnecting different simulation tools to solve such problems is hardly applicable as incompatible file formats, mark-up languages and models describing simulated objects cause an inconsistency of data and interfaces. This paper presents the architecture of a framework for adaptive data integration that enables the interconnection of such numerical simulation tools of a specific domain.