Since the Nineties, a constantly growing number of digital assets concerning Built Environment have been produced worldwide by diverse stakeholders from the fields of Architecture practices and engineering contractors, of the academic research in History of Architecture and in Archaeology, of e-Learning, of Heritage Governance and of Building Control. Reflecting the aims, the scopes and the policies of their producers, these assets span a wide variety of digital formats including 2D vectorial drawings, 3D meshes, raster images, audiovisuals, raw points clouds and GIS data-sets.

Collections of these digital records are spread over repositories with relevant differences in scale, policies and functions, ranging from the documentation held by Architecture practices to the records deposited in the national legal repositories. But, the smallest collections are unlikely to implement consistent curatorial policies, thus, the digital curation life-cycle of these digital assets often begins with their ingestion in institutional repositories. Since the vast majority of these assets is represented by documentation held by Architecture practices and other assimilable stakeholders which is neither intended for ingestion in institutional repositories nor for long term retention in commercial archives, the risk of data loss by complete lack of curation actions results affecting the wider section of this records’ category.

Further, the heterogeneity and complexity of these digital objects, as well as the prevailing role of major repositories in establishing preservation and curation policies, cause a lack of specificity in their digital curation that could endanger the persistence of information, damaging culturally and economically their stakeholders. Indeed, the results of studies concerned with the curation of CAD files, demonstrate these problems:

• the appraisal, selection and ingestion of these digital objects do not rely on the needs and expertise of their users' community, and in doing so the future accessibility of valuable documents is seriously endangered;
• the preservation and storage actions adopted by the repositories similarly do not consider the needs of their stakeholders since the original framework of data production tend to be obliterated;
• the access models and the browsing facilities provided by repositories are not viable to promote either their professional reuse or the aware management of their intellectual rights.

Nevertheless, in the last ten years many international initiatives have been dedicated to both identify and implement preservation and curation policies especially for the digital documentation of cultural heritage, starting from the seminal NINCH Guide to Good Practice in the Digital Representation and Management of Cultural Heritage Materials to the latest 3D-COFORM project. Providing effective preservation tools and curatorial frameworks for the harmonization, coordination and the accessibility of digital assets, these projects have demonstrated the necessity of basing the curatorial policies of a collection on the needs of their users as much as on their integration with other contexts.

The metadata annotations and the semantic browsing of these digital objects are a relevant part of the strategies to enhance their longevity, supporting their reuse, dissemination and automated preservation. Although being fostered by internationally recommended ontologies allowing horizontal specifications of data annotations, most notably by the CIDOC-CRM, the implementation of semantics for digital Built Heritage are still either undeveloped or isolated. Through the recent publication of the London Charter within the EPOCH research program, a major advance has occurred in the fields of descriptive and structural metadata for 3D documentation of Built Heritage but, although supporting the spreading of a trusted digital scholarly communication, did not affect either the accessibility or the preservation of these digital objects.