The processes described in the use cases shall overcome the existing challenges through a number of objectives. DURAARK addresses the following scientific and technical objectives:
Workflows and streamlined processes for the long term archival of the broad spectrum of architectural data. Workflows and tools for OAIS compliant ingest, management and delivery of the various types of architectural data ranging from low-level point cloud scans over legacy 3D CAD models up to highly enriched Building Information Modeling (BIM) will be set up and developed. Furthermore a high focus will be put on pre-ingst tasks, addressing problems as early as possible in the curation and preservation chain and enabling the various stakeholders to make informed decisions about the sustainability of their data.
Semantic enrichment with metadata. Vocabularies/ontologies will be specified and developed. This process will be based on an alignment with already existing metadata schemes tailored to 3D modelling of architecture and on an extension to include key concepts that ensure the interoperability between 3D modelling tools and distinct 3D model libraries and archives. Furthermore enrichment techniques will be applied which allow for the enrichment of naturally poorly structured 3D model metadata with additional knowledge from the LOD cloud.
Consistent naming schemes and ontologies for architectural entities. Semantically meaningful information of buildings and their components will be captured by employing newly developed methods and technologies for the creation, long term persistence and management of ontological concepts and their facets.
Furthermore, approaches for distributed collaborative authoring of the ontological content, its approval and re-visioning for long term access will be devised.
Render semantically low-level 3D architectural data meaningful and interpretable. Unstructured 3D point clouds and legacy 3D CAD data with no semantics will be lifted from a purely geometric level to one that contains textual and structural information which is meaningful to the architectural domain and facilitates the understanding and semantically meaningful retrieval of the content in the near and distant future.
Dealing with point cloud data. The extraordinary size of point cloud scans renders an interactive ingest, management and delivery of raw data practically impossible. As this problem did not occur for traditional 3D architectural data such as low level CAD objects, it poses a new challenge. Novel techniques for the interactive retrieval and delivery of fragments of interest as well as special compression techniques suited for the long term archival of point data will be developed.
Documenting the changing state of built architecture. Strategies and methods to link as-is building representations in 3D point clouds to either previous point cloud representations of the same building or to the original architectural CAD files will be researched and developed. This linking will provide means for the long-term documentation of a building’s changing state.
Storage of highly enriched 3D architectural data in an existing OAIS compliant long-term preservation system. The development of a submission application for the highly enriched data as well as the development of a workflow within the long term archival system Rosetta to handle the highly enriched data from ingest to storage and from storage to access will enable the archiving and accessibility of complex architectural 3D models. As Rosetta is an OAIS-compliant digital preservation system used by several heritage institutions world-wide, the developed tools will be usable within the digital preservation community. Furthermore, a preservation planning concept which collects all information from a technical as well as business perspective for the enriched architectural data will be devised and established.