The intensive use of computational tools like Computational Fluid Dynamics and Finite Element Methods in development, test, manufacturing, and service has resulted in a tremendous increase of data that is managed in an engineering context.
The integration and combined analysis of data from multiple disciplines resulting from cyber physical systems potentially allows for a drastic increase in efficiency and quality during the overall product lifecycle.
In order to optimally utilize this data, algorithms are developed which are able to autonomously identify on the one hand system states, operation patterns and eventual unexpected system behavior. On the other hand to identify and explain system behavior and interaction with the environment to improve and to innovate the system during current and future design phases.
Complex dependencies are not necessarily obvious during the design phase, especially for interacting systems in a dynamic environment. In Data Analytics, a diverse set of methods from statistics and computational intelligence is applied to enrich the knowledge about the system and its interaction with the environment including the user, to maintain and to improve the systems.
Data Analysis for Industrial and Engineering Design Data
Modern engineering development cycles are driven by CAE methods and the computational analysis of digital simulation models, leading to an ever increasing availability of complex, large-scale data sets. To handle this complexity, data-driven approaches from machine learning and artificial intelligence are explored to provide novel insights into engineering processes.
We place special emphasis on the integration of shape and structural information by deploying and researching machine learning techniques for processing of engineering data, in particular 3D designs. For dependency analysis, nonlinear multivariate methods are deployed, providing additional insights into simulation results. Interaction graphs are created to illustrate these newly found interrelations between spatially decoupled surface areas.
The ability and analysis of relating surface and shape information to performance indices, assists the engineer in the exploration of the design space and significantly reduces development time.