Call for Papers
Daniele VILLA | Lorenzo CECCON
(Politecnico di Milano, Italy)
Keywords: AI, HBIM, Remote Sensing, Diagnostics, Ontologies
Call: AI shows great potential in helping formulate H-BIM model hypotheses of surveyed heritage sites. It can in fact link together, on the one hand, survey data from a specific site and, on the other hand, wide data-sets of other already analysed heritage sites: raw survey data and corresponding validated reconstructions as shared in common repositories. Based on such data – structured according to ontologies that embed previous research results as wrapped into the H-BIM model – AI could “make sense” of the survey data at stake and help build the relevant H-BIM model embedding such semantic taxonomy. It would then “suggest” reconstruction hypotheses, both as regards the inner structure and materiality of the surveyed sites, such as the wall material composition layers, and the historical/philological development thereof, such as the construction phases and “styles” of the site.
The session will highlight the current uses of AI for Cultural Heritage, including, but not limited to, point cloud segmentation, bimification, model-fitting techniques and fragment re-composition. It will invite scholars to devise a new paradigm in the use of AI to create historical “Digital-Twins”, to be progressively fit to the survey and historical findings over-time. The H-BIM model would then not just represent a final reconstruction result, rather help in the process of hypothesizing and testing reconstruction hypotheses, including the site diachronic development and decay history. Special attention shall be thus dedicated to the structuring of Open Data Repositories and of the sematic HBIM families contained therein, as a key “bridging” knowledge-base between survey data and H-BIM modelling, i.e. the taxonomic “glasses” through which AI would see the world. The aim is encouraging the research community to create a sound common process and knowledge build up to “fuel” and exploit AI tools.
Submission (open April 15, 2020)
Mind the guidelines