DURAARK

A European research collaboration

Problems and project
DURAARK is a three year research project, which tackles the challenge of developing secure and efficient long-term archival processes for 3D architectural data. The project has a focus on the combination of unconstructed data from Building-Information-Modelling (BIM) environments and will develop strategies to search and detect architecturally meaningful information in the automatically created rich data in order to overcome current problems that inhibit a successful long-term preservation of 3D architectural content.

Problems to tackle:
Non-existent semantic relations between architectural representations in Building Information Models, point clouds and geo-referenced data from web and images.

No automated systems to detect and search for architectural structures in point clouds.

Heterogeneous and inconsistent metadata schemes and ontologies for the description of building elements and their properties in highly enriched BIM models.

Fast changing development of architectural software, files and referenced web-based data as metadata schemes prohibits successful and easy long-term archival for companies and institutions.

Aim and approach
The DURAARK project will develop query and preservation tools that are tailored to the domain of architectural 3D content. The objective is to ensure future-proof reuse of 3D objects, using open file standards and providing techniques to gain new architectural information through enrichment and automated combination of heterogeneous data.

The research will be developed, implemented and evaluated with partners from the professional communities in architecture, engineering and building administration. Cases and short-term hands-on projects will help to guide and evaluate the overall project.

DURAARK will develop:
Methods for a metadata rich long-term archival of architectural content with its vast domain-specific spectrum of semantic detail levels and fragmented nature of naming schemes and dynamic references to other web-based content.

Tools to automatically link point cloud scans and BIM data and the enrichment of architectural data with other georeferenced content as mages and semantic information from the web.

Queries on point clouds and BIM data in order to allow a future long-term archive to answer questions as they appear in a practice that needs to work seamlessly across heterogeneous sets of data.

Tools for the automatic detection of architectural meaningful structures as walls, openings and floors in point clouds.

Means to allow queries about the changing state to buildings by accessing data from different periods of time.