The steady growth in size and complexity of simulation data, as well as the way it is postprocessed, in many situations does not comply with the requirements of efficient visual analysis. This becomes obvious when data from very large simulations of, e.g., airflow around vehicles, is maintained at an automobile manufacturer‘s headquarters, and engineers at remote locations are working with the data in a loop of visual analysis and model modification.
Today’s visualization processes like remote visualization or the bulk transmission of all data before visualization do not scale with data growth and increasing numbers of concurrent users. This project aims at developing novel data representation and transfer methods to integrate into in a new visualization pipeline, which will be capable of handling and visualizing efficiently increasing magnitudes of remote data.
To these ends, hierarchical compression of irregular grids is combined with progressive on-demand data transmission and local volume rendering of tetrahedral multiresolution meshes. The interface to the data is provided as a generic layer that models general graphs and grids, which enables zero-overhead integration with distributed caching services as well as existing simulation and visualization systems.
Contact: Prof. Dr. André Hinkenjann, Prof. Dr. Andreas Priesnitz, Oliver Jato, M.Sc., Thorsten Roth, M.Sc.
Partners: SIDACT GmbH
Duration: June 2018 – September 2020
This project is supported by the Federal Ministry for Economic Affairs and Energy (BMWi) under the Central Innovation Programme for Small and Medium Sized Enterprises (ZIM) grant No. ZF4120903MC8.
Subproject of the funding proposal HACS: Hierarchical Methods for the Efficient Analysis of Cloud-Based Simulation Data.